Protection from Online Falsehoods and Manipulation Bill

Steven Barnett is a prominent parliamentary advisor and Professor of Communications in the UK. He said, "In the lead up to the June 23 European Union referendum, British mainstream media failed spectacularly. Led, inevitably, by the viscerally anti-EU Daily Mail, Sun, Daily Express, Telegraph, most of Britain's national press indulged in little more than a catalog of distortions, half-truths and outright lies. It was a ferocious propaganda campaign in which facts, sober analysis sacrificed to the ideologically driven objectives of editors and other proprietors."

The Pro-Leave Camp in UK – The Daily Mail, The Daily Express, The Sun, The Daily Telegraph – combined readership of 28 million, used sensational headlines; outright lies. Sunday Express said "12 million Turks would move to the UK" if Turkey joined the EU; a complete falsehood. Later, it admitted it was inaccurate

Government on the one side; people on the other; media is intermediary. If the media regularly trade in lies to attack the Government without basis, this happens – trust in Government goes down. Institutions will be severely damaged.

And the media can destroy institutions. An example of a great institution that is being savaged is the UK Courts, among the best in the world. I have spoken about this in the House previously. I do not wish to go into it but I have, in the attachment, set out how the British media have severely damaged the British judiciary. [Please refer to Annex 1.]

Traditional media holds power over society's information. It has the ability to influence minds and viewpoints. When media acts responsibly, it serves democracy. When they do not, it damages democracy.

So, in many countries, traditional media has played a big role, in the loss of trust.

I spoke about inequality, inequity, political systems not delivering results, media abusing its power, new media being abused. I have sketched out the first three. New media is the subject of this Bill. I will come back to it and deal with it in greater detail.

But let me first complete the larger point on the consequences of the four aspects coming together. Members can then see what happens when you do not deal with the very real dangers of these developments.

In an active democracy, the foundations include: trust, free speech and the infrastructure of fact. The four elements I referred to, have combined, like a battering ram, to damage, destroy these foundations.

When people lose trust, when they lose faith, when there is no proper public discourse, when infrastructure of fact is damaged, then democracy, societies are at serious risk.

Populism will then rise. Violence will rise, particularly towards the minorities, the weak. The ability of countries to face challenges will be weakened. The crisis of trust in many countries has opened wide the doors to dangerous, destructive politics, populism. Populism both exploits and deepens the loss of faith in the system, making it harder for institutions to correct, and to find solutions.

Gallup has said, and I quote, when people hold "low trust in government and low or static expectations for their future lives", support for populist, anti-establishment politics increases. A politics that can – and has, in some countries – destroy democracy and replace it with corruption.

The 2017 Edelman Trust Barometer drew a direct link between the lack of trust in public institutions and these populist movements. Forty-four percent of those who voted to leave the EU believed the system was failing, and held fears about at least one major societal issue, compared with 20% of those who wanted to remain in the EU.

Populists used these fears as ammunition to feed the crisis of trust in institutions. And 2016 saw populist movements make major gains. The Eurasia Group said populism is a force in US politics. And in Europe, populist political parties are getting close to or are in government positions in several countries. Populism also reigns in some Latin American countries and has gained footholds in Asia.

And the assessment by the Eurasia Group is populism is "likely to intensify and spread over the coming decade, weakening governments and de-legitimising political leaders as a consequence."

The result will be a set of unyielding, centrifugal forces, creating a global environment described by the Eurasia Group as "the most dangerous it it has been in decades"

This is the serious situation that faces many countries, and we will face it as well. That is why I referred to Gotterdammerung earlier.

Many issues: inequality, inequity, political system, failing to deliver, traditional media, new media, the impact leading to loss of trust and to populism. We have to avoid this trajectory, which means we have to deal with each of these issues.

This Bill is an attempt to deal with one part of the problem: The serious problems arising from falsehoods spread through new media. And to try and help support the infrastructure of fact and promote honest speech in public discourse. It is an important part, even as we work on the other aspects.

Now, I turn to falsehoods and new media. Many societies, including ours, are pluralistic and diverse, with competing interests and identities, with religious diversity as well. Various interest groups – NGOs and other associations are there. Each pursues its interests, causes and passions. Such diversity is never easy to navigate. It can give rise to instability and conflict.

James Madison, one of America's founding fathers, argues that the "factions" in a diverse society cannot be controlled. Only the effects of such factions can be countered. Decision-making with diversity is not always easy. And a key foundation of a democracy is public discourse.

Ideally, public discourse will help citizens understand complex policy issues. It will guide policy-makers to make optimal decisions. It will shape differing viewpoints and expand common ground.

But public discourse can only take place when there is free and responsible speech.

The pre-requisites for national conversations are a common vocabulary, an underpinning of facts and that provides a platform for accommodation and compromise amongst diverse voices in society.

A critical piece of infrastructure in these conversations is fact, and the infrastructure of fact.

Like public infrastructure, society depends on it. It provides society with a shared reality. This is necessary so that we can have diversity without conflict, and public participation, while still getting decisions made. Without it, our political system will malfunction.

William Davies, the English sociologist and political economist has said, and I quote, "The fact that millions of people are able to believe the same things about reality is a remarkable achievement, but one that is more fragile than is often recognised."

A critical reason for this remarkable achievement, and I quote Davies again, is: "something so ubiquitous, so ordinary, that we scarcely ever stop to notice it: trust.”

This is the trust that when public institutions – the Government, the media and other institutions – share a piece of information with the public, they do so honestly.

Trust in public institutions is a cornerstone of the infrastructure of fact. The belief in the authenticity of the source, whether it be the media, the Government, experts or other authoritative sources, makes society accept facts. Authenticity, of course, has to be earned, maintained. And if an usually trusted source is not telling the truth, there must be avenues to expose and there must be consequences.

So that rigour helps in maintaining authenticity. People will then believe, if it said, it is likely to be true. And if not true, it will be exposed.

These are amongst the foundations on which modern societies are based. They are bigger than the Government of the day. They are bigger than any political party. This is about the basic structure of society

New media has been heavily exploited to batter this infrastructure of fact which in turn, weakens trust in public discourse, in institutions, in democracy itself.

I will now set out how falsehoods have been weaponised to attack the infrastructure of fact, to destroy trust, and to attack societies.

New media is an information super highway. It has got many by ways, links to different groups in society, and to everyone. It has now been used to send out on an industrial scale, falsehoods, to mislead people. Broad sections targeted, but also specific groups targeted, and individuals are targeted.

The attack using falsehoods on social media comes from several sources: (a) foreign countries using information warfare; (b) profit-driven actors; (c) deliberate actors, for political ends; and (d) people with prejudices, seeking to harm other groups. The Select Committee noted this. So, let us look at some sources of falsehoods.

First, foreign countries. There is a military doctrine that has been developed for the Internet age. General Gerasimov, Russia’s military Chief of Staff, there is a doctrine named after him called the Gerasimov Doctrine. Basically, he says, the "rules of war" have changed. Non-military measures, including information operations, they can be harnessed to, in turn harness what he calls the "protest potential of the population". What does it mean? Information operations can target and create internal opposition as a "permanently operating front" throughout the target country. These non-kinetic military measures, in many cases, can exceed, as his doctrine says, the power of force and weapons. Even though military or overt violent measures are not being used, the target states’ national sovereignty and security are threatened and violated. In this way, the lines between war and peace have now blurred and wars no longer have to be declared.

Security experts gave evidence to the Select Committee. Dr Shashi Jayakumar said that in modern information warfare, "seeding internal opposition within the target country is extremely important." And he said that "technological tsunami" has given aggressor states the ability to "subvert individual slices" of the target country in a manner unthinkable just two to three decades ago. A national security expert from Latvia, Dr Janis Berzins, said that the notion of a broken social contract is the main vulnerability exploited by foreign adversaries. In his words: "It is easier for the adversary to achieve its objectives if the society of the state being attacked believes that their country is a failed state that does not care for the interests and needs of the population, and the loss of current statehood will bring better living conditions."

Singapore is a specific and vulnerable target for some very precise reasons. It has military superiority in this region – experts said this. That superiority in conventional military forces means it will be futile to start a war with Singapore. And, therefore, militarily weaker countries will then focus on other means to weaken Singapore, sap our will from inside, create deep internal divisions and keep us in a permanent state of internal dissension.

The evidence is that this is already happening. The Select Committee heard it. We know it is happening even though we do not come out in public and say it very openly. It is happening to sap people’s support for the SAF, for defence, to try and shift Singapore’s foreign policy as well. These are not issues that we should dismiss lightly.

Just assuming we go on the basis of the evidence that was presented at the Select Committee, it has happened elsewhere and I will give you some evidence of how it has happened. Take Ukraine. A foreign country, which the Select Committee did not really name, used falsehoods to build a narrative that the Ukrainian government was fascist and corrupt. It spread online falsehoods about atrocities being carried out against a particular community in Ukraine. For example, that Ukrainian soldiers had crucified a child, which was later debunked. Volunteers who fought against Ukraine said they were motivated because of these supposed atrocities. Consequences? Loss of sovereignty, part of territory, loss of lives.

If you take the Czech Republic, disinformation operation by a foreign country was used to turn domestic sentiments in favour of a foreign state’s geopolitical goals. One objective was to make people believe that the US was responsible for the influx of Syrian refugees into Europe and the conflict in Ukraine. The disinformation appears to have had some impact. In 2016, 50.2% of Czechs believed that the US was responsible for Syrian refugees coming to Europe, 38% believed that the Ukrainian crisis was caused by the US and NATO. So, it was done to weaken Czech support for NATO and Ukraine.

Take Germany. If it can happen in the UK and if it can happen in Germany, I believe it can easily happen here. A girl fabricated a claim that she had been assaulted by three Middle Eastern migrants. Foreign media outlets reported on that widely, suggesting it was true, specifically from one country. Reports were then spread on social media. The Berlin authorities investigated, confirmed the girl’s claim had been fabricated. But falsehoods were circulated online that the police had failed to follow up on the case. Thousands demonstrated on the streets to "expose" the government’s attempts to cover up the crimes perpetrated by the refugees. That same year, a far-right populist party made unprecedented gains in the regional elections and most of its support came from the same minority constituency as the girl who made the rape claims.

Sweden. A Swedish defence agency said false information about subjects, such as NATO, immigration and terrorism, are spread "on a daily basis in Sweden."

In many of these countries, trust is being eroded in institutions, populism is on the rise. Since Russia’s name had come up in the media, the Select Committee invited Russia to help us and the Russian Embassy very helpfully sent us a letter that Russia is also very concerned about fake news and has been a victim of fake news. So, we have no doubt that Russia takes this matter very seriously.

Next, apart from foreign state actors, commercial profit is another dimension. Digital advertising models have turned websites into virtual real estate. So, value depends on the attention attracted, with every click, every view, digital ad revenue is earned. This business model has created an attention economy, content that stokes fear and anger and good for attracting attention. Falsehoods can help people earn large sums of money and have political impact.

In the US, an American named Paul Horner set up at least 20 fake news websites. Some used deceptive URLs. It tricked readers into thinking that they were mainstream sources like ABC News or CNN. He was prolific. He used his websites to publish fake stories, stoke partisan engagement, claimed that they were satirical, said he assumed people would fact-check. But many did not, and they were fooled. Some examples of his falsehoods: during a Government shutdown, President Obama used his own money to keep open a government-funded Muslim culture museum. Fox News reported the story as a fact, before retracting it. False article: protestors were paid to protest against Mr Trump; re-tweeted by the Trump campaign; 20 million Amish people had committed to vote for Mr Trump, turned up in Google News, had 750,000 page views in two days. He made $10,000 a month from Google AdSense. He was opposed to President Trump but targeted conservatives with his fake stories because he found it more profitable. A BuzzFeed news investigation found a network of at least 43 websites which together published more than 750 fake news articles – a small study. All of them used Google AdSense to earn digital advertising revenue.

Towards the end of the 2016 US presidential elections, a small town in Macedonia became well-known as the registered home of at least 100 pro-Trump websites filled with sensational falsehoods. Macedonians experimented with different target audiences. Fake stories could gain the most traction amongst Trump supporters, they found. Sometimes they wrote their own articles, but often would just reuse falsehoods from alt-right websites in the US. In the four months leading up to the elections, one young Macedonian earned nearly US$16,000 from the two websites he ran. For five or six hours of work he could earn about $1,000 a month. So, as Samanth Subramanian, a reporter with Wired, said, "This is the… disturbing heart of the affair: that the Internet made it so simple for these young men to finance their material whims and that their actions helped deliver such momentous consequences."

Third category, deliberate individual actors. You had foreign state actors; you had people doing it for money; now, people who do it deliberately, and groups. The Select Committee found that in several countries, both local and foreign civilians had spread falsehoods for political causes, affecting both their own countries and other countries as well. Example again – far-right, nationalist groups, promoting more extreme politics. Domestic alt-right in the US used falsehoods, drove major false narratives during the 2016 US presidential elections. More recently, in 2018, in the Brazilian presidential elections, local far-right groups coordinated a network of fake social media accounts, spread misinformation in support of right-wing candidates.

Populists use lies to attack institutions, invoke divisive rhetoric. They use conspiracy theories to explain complex issues in simple terms and trying to make people believe them.

Truth then becomes completely irrelevant. Even the most extreme lies which we might think people will normally dismiss, become believed and it impacts very badly on public life.

British historian and journalist Anne Applebaum said this about populist movements: "They don’t require belief in a full-blown ideology...most of them don’t deploy propaganda that conflicts with everyday reality. And yet…all of them encourage their followers to engage, at least part of the time, with an alternative reality."

In the UK, falsehoods were spread extensively during the EU Referendum. Immigration was a key issue. I spoke earlier about false claims in traditional media about Turks, Turkey and Brexit. Digital advertisements were also run by the Leave EU Campaign that Turkey was joining the EU, 12 million Turks would in that event move to the UK, if it remained in the EU.

Foreign interference was also at play. Fake foreign-linked accounts posted more than 45,000 messages about Brexit in the 48 hours during the referendum. A large part of the foreign-linked content related to refugees and immigration. False stories. Stories of illegal migrants attacking women, concocted. But these falsehoods were used to create an alternative reality. If you believed them, you would believe there was a conspiracy by the ruling elite to turn the UK and Europe into a Muslim caliphate. Muslims were campaigning for Sharia law to govern the UK, the Mayor of London, a Muslim, was sponsoring them and, in the UK, there are areas where Sharia law dominates and non-Muslims cannot enter. Sounds outlandish. Who will believe? The British are very sensible people. But even the British fell for it.

In 2018, YouGov did a survey of over 10,000 people, so it was a substantial survey. Thirty-two percent believed the falsehood about "no-go" areas under Sharia rule in the UK. Forty-nine percent of those who voted to leave the EU stated that this was true. It also created a permissive environment for hate. In the month after the referendum, there was a 41% spike in hate crimes and the majority of crimes were motivated by race, including crimes against migrants.

Brexit altered the course of British history. It was one of the most important events in their recent history. A massive falsehood campaign may well have affected the outcome. An investigative report in The Guardian by a British journalist, Caroline Cadwalladr explained. She went to a town in Wales that probably received more EU funds than any other town and which still voted to leave the EU. Sixty-two percent of them voted to leave. In this town, the EU was funding a 350 million pound regeneration project, a 33 million pound college for further education, which had 29,000 apprenticeships for young people to learn a trade, a 77 million pound road improvement scheme and a 30 million pound railway line. The town had one of the lowest immigration rates in the country. It was in Wales, which was a net EU beneficiary.

But when Caroline interviewed the people in the town, she discovered the existence of an alternative reality. People believed the town was sending more money to the EU than it was receiving. The EU had not only done nothing for the town, it had also brought a huge immigration problem, including from Turkey. The Welsh town had been a left-wing Labour stronghold. Yet, the people were repeating information usually found in right-wing newspapers. She discovered that they were getting their information or misinformation from targeted digital ads on Facebook.

One of the most controversial claims in the UK during the referendum was that the UK sends 350 million pounds a week to the EU. Not only plastered on buses but also run as digital ads aimed at specific groups of voters. The UK Statistics Authority said the figure was likely closer to 136 million pounds per week. Despite the corrections, a 2018 King’s College London study found 42% of the people who had heard this claim believed it to be true and 22% were unsure. Only 30% correctly believed that it was false. For a referendum that may have irreversibly altered the course of British history, these are serious consequences.

In the US as well, falsehoods and lies have been spread on a massive scale. A study from Stanford University shows that, on average, each American adult read about three false stories in the months leading up to the 2016 US elections. And the researchers that compiled the fake stories that had been debunked by fact-checking websites found that these falsehoods were shared about 38 million times, leading to around 760 million engagements with the content. And the false narratives were driven by the people who identified with the home-grown alternate-right movement.

Largely sought to denigrate the establishment, attack Mrs Clinton’s campaign. Also at play: sophisticated foreign information campaign that sought to influence the outcome; undermine democratic institutions and the democratic ideals. There was the conspiracy theory that Mrs Clinton and other top Democrats were part of a child paedophilia ring operating out of a Pizza restaurant in Washington, DC. Sounds completely crazy. Utterly unbelievable if you think about it. The rumours began on Twitter, spread to other websites and online forums. Claims got louder, hacked emails from Mrs Clinton’s campaign were distorted to support the conspiracy theory. An American man showed up at the Pizza restaurant to “self-investigate” and he brought a gun along with him. People demonstrated in front of the White House declaring that the theory was real.

The falsehoods helped to create an alternate reality, one where a "deep state" existed within US institutions, which was conspiring against the American people and, if you are a part of this echo chamber, you would be told that then-President Obama was colluding with the UK spy agency to spy on Mr Trump. Mrs Clinton was linked to a mysterious explosion that killed one of her employees. And the 2018 Florida school shooting was really a secret government operation. Foreign agents infiltrated, exploited this alt-right movement using fake social media accounts. They pretended to be real Americans, amplified the falsehoods that originated from these websites, targeted echo chambers with claims that Mrs Clinton wanted to apply Sharia law in the US, that Mrs Clinton was actively arming the terrorist group ISIS. The foreign agents also infiltrated other social media movements to spread falsehoods on both sides of issues – they were agnostic about that – and to create more divisions. They amplified the falsehoods, they widened the divides. So, the political ground became fragile, fraught and conducive for foreign agents to operate.

So, the experiences of these two countries show how, through a combination of falsehoods and digital technology, the foundations of democratic society are severely attacked. Falsehoods are used to undermine public trust which is the cornerstone of our infrastructure of fact. They are used to divide and polarise, tearing the social fabric. And democratic discourse, accommodation and compromise become very difficult. In these conditions, the political centre becomes hollowed out and people are driven to extremes.

If you take France, the Yellow Vest movement, at the centre of the movement are these "Anger Groups" on Facebook. They are online hubs for falsehoods. They appeared almost a year before the street protests began in Paris. In the lead-up to the protests, falsehoods were used to increase the sense that the system was failing and turning against the people. Various falsehoods – that the French Constitution had been nullified in 2016 by the then-Prime Minister; that President Macron was going to sign France’s sovereignty away at the UN conference; a million Germans had protested increased fuel prices, that President Macron wrote to Paris police to use force against the protestors and so on. And during the protests, misinformation in Yellow Vest Facebook groups and pages reached over 105 million views and four million shares in five months. Images were falsely captioned as bleeding Yellow Vest protestors and then they were used to make the claim that media and government were hiding police brutality and violence against these protestors.

During the 2016 US Presidential Election, disinformation came from foreign operatives. But in the mid-term elections, such campaigns were more domestic. Americans were targeting other Americans, using the same strategies that are said to have been introduced by the Russians. Sometimes, activists worked with foreign states. The Select Committee Report has pointed out that state actors have formally coordinated with, co-opted other private-sector actors, including private industry, civil society organisations, fringe movements, volunteers who ideologically support their cause. The Select Committee Report also mentioned an Asian country, it did not name it. It is said to have an online cyber army and volunteers who promote the government’s policies and attack those who criticise those policies.

Social media has also enabled hate to thrive. Falsehoods are often centrepieces of hate propaganda and, over the past few years, people have used online falsehoods to promote anti-immigrant, anti-Muslim prejudice. After the terrorist attack in Paris in 2015, a video was posted, described as showing "Moderate Muslims" celebrating the attack. It was actually a video of people celebrating a cricket match victory in Pakistan. But nearly 500,000 views within a few hours. Video was spread again after the terrorist attack in Paris in 2017 with the same false caption.

The terror attack at Westminster in London in 2017: it became viral. It is a photograph of a Muslim woman walking past victims. She was on her phone, face hidden from view, was falsely accused of ignoring the victims and treating the attack casually. Later, it was later discovered that the tweet originated from a fake foreign account.

When moderate Muslims demonstrated against terrorism in London, a false story was spread that the demonstration was staged. And in April this year, falsehoods were used to turn the Notre Dame fire into an anti-Muslim narrative. A fabricated quote, ascribed to a Muslim US senator saying "they reap what they sow". A video of Notre Dame burning with shouts of "Allahu Akbar" edited over.

In Indonesia, a "Muslim Cyber Army" used falsehoods and hate speech to inflame sentiments against gays, Chinese. We have also seen this in Germany, Italy and Brazil and all have rising populism, no coincidence. The psychological evidence is that mere exposure to conspiracy theories, even if they are dismissed, makes people less likely to accept official information or engage in politics.

Conspiracy theories harm trust in institutions overall, not just the specific institutions that they relate to. And, so, people mistrust the very existence of an infrastructure of fact and disengage from public discourse altogether. So, as William Davies said, "…when trust sinks beneath a certain point, many people may come to view the entire spectacle of politics and public life as a sham…"

So in the digital age, almost anyone can make a falsehood go viral, or run a disinformation campaign. It has happened in Sri Lanka. In India, child abduction rumours spread on WhatsApp – 69 mob attacks, 33 deaths. In Mexico, same child abduction rumours, horrific lynching of two men. This year, in France, child abduction rumours targeting the Roma people led to several violent attacks on the Roma.

Harm to public health when falsehoods are spread about health care. Financial markets can be affected very quickly. A false tweet that the White House had been bombed led to a massive fall in the stock market. So, examples are innumerable. In every country, this is happening. No one can disagree that this is a serious threat and that it has to be dealt with head on.

I will say this to Members of this House on both sides and the Nominated Members of Parliament that even when this happens in Singapore, there will be a deep damage to the institutions which are beyond and above politics. We are seeing this happen before our very eyes in other countries which normally you could have assumed to be stable, much more stable than Singapore, much bigger than Singapore. And if that happens, that will damage society beyond repair. So, there is no benefit to anyone, regardless of any political persuasion. There is no benefit to see this happen because there is no benefit, whether political or otherwise, in seeing this happen. For anyone.

What are the tools that such actors use and what is the cost of such tools? Digital technology has given falsehoods a new power. The Select Committee found "…considerable evidence was given showing how modern digital technology has made the creation and dissemination of falsehoods easier, cheaper, more profitable, transforming it."

I will mention three tools: fake accounts; digital advertising; and algorithms used by platforms to rank content.

Fake accounts have been described as "foot soldiers" of disinformation. They may be run either by humans known as "trolls", or may be automated, in which case, they are called "bots". Fake social media accounts manufactured to manipulate. Some of them cultivate persuasive online personas, gain followers, both real and fake, and used as fictitious leaders of public opinion, using falsehoods to sway minds, create impressions of public sentiment. Bots are used to artificially amplify falsehoods – megaphone for falsehoods. They draw attention to falsehoods, affirm them and make them appear to be more believable.

Digital advertising tools are used to target falsehoods at susceptible segments of the population. Search engine results manipulated so that false articles are prioritised. Online echo chambers are created and exploited. Social media groups and online chats are used to polarise and mobilise people to orchestrate conflict. Falsehoods today travel easily and widely across multiple platforms. In 2017, a news article that a Russian aircraft had managed to electronically disable a US warship was posted on a Russian state-controlled news site. The article used a parody that had been published a few years before and presented it as truth. You can see how it is not just the labels that matter. People can be made to believe that parody is, in fact, true. So, you need to look at the material objectively.

So, this slide, designed to glorify Russia, undermine confidence in the US Navy. In two days, it was picked up by mainstream outlets in the US and Europe. You can see how it spread. [Please refer to Annex 2.] In less than seven days, cross-posted on dozens of news sites and their social media pages, catered to different countries and demographics. Fox News’ version was shared over 27,000 times. The Sun’s version was shared over 10,000 times. Some of the more prominent examples over the last three years.

In 2016 Brexit: research suggests that bots were generating up to 20% of Brexit-related tweets in the months leading to the Referendum, including anti-Muslim falsehoods.

In the 2016 US Presidential Election, a foreign troll factory conducted a disinformation campaign using 50,000 bot accounts, over 3,800 Twitter accounts and at least 470 fake Facebook accounts. Facebook’s best estimate from 2015 to 2017, approximately 126 million people may have received content from accounts associated with this troll factory.

In the 2017 French Presidential Elections, emails from then-candidate Emmanuel Macron’s campaign were hacked and leaked. So, #Macronleaks hashtag was used to guide Twitter users to false claims that the leaked emails showed evidence of illegal activity by Macron. Hashtag was then amplified through a network of trolls and bots and reached 47,000 tweets in less than four hours after the initial tweet.

"Coordinated inauthentic behaviour" by fake accounts was also seen in the US' 2018 mid-term elections; 2018 Brazilian Presidential Elections; the US anti-vaccine debate; and in Indonesia.

Fake accounts have been used in very sophisticated ways. A Twitter account belonging to one "Jenna Abrams" was created in 2014. This appeared to belong to a young American woman. Once she had built a following, she pushed a divisive set of views on immigration and Mr Trump, especially closer to the 2016 Elections. At one point, she had over 70,000 followers, quoted by dozens of high-profile media outlets, including The Washington Post, BBC and the New York Times. But it turned out to be a fake account created by a foreign troll agency.

The Twitter account impersonating the Tennessee Republican Party, also created by the same agency, spread falsehoods: President Obama had admitted he was Muslim; an employee of Clinton’s was killed in a mysterious explosion in Washington and Clinton was involved. At one point, it had over 152,000 followers. The real Tennessee Republican Party account had less than 14,000 followers.

So, trolls, bots, can and have been used in Singapore. We have noticed these spikes in activity from inauthentic accounts when we have discussions on various issues of public concern. One example, whenever there is a bilateral issue with Malaysia, these go up. We recently estimated that it went up by 30%. Such activity creates alternate realities. It manipulates perception, creates the impression that there are many voices, shouts down other viewpoints through fake accounts, shifts public opinion, erodes trust and undermines institutions.

If you turn to digital advertising, it is done to sway public opinion used on both sides of the Atlantic. Foreign operatives were using it in a disinformation operation to influence the US Elections. It was also used by the Vote Leave campaign in the UK. These tools allow messages to be micro-targeted with a high degree of precision at specific groups based on a variety of indicators, including people’s fears and prejudices. They are able to do so because the personal data of each user is harvested by platforms like Facebook. In the US, foreign operatives used US$100,000 to spread Facebook advertisements to 126 million Americans. It is inexpensive compared with traditional advertising. Almost 2,000 ads used interest-based targeting and, of those, 800 were geographically targeted, including at swing states. Remember, in some swing states, the margin of victory was in the tens of thousands.

In the UK, the Vote Leave campaign spent more than 2.7 million pounds on targeted advertisements. Targeted digital advertising was used in these campaigns because it was effective. Study by network theorists showed when falsehoods are initially aimed at those predisposed to believe them, they spread further. Another troubling aspect of targeted advertising is that it is hidden from public view. Others cannot see the falsehoods being spread, they are unable to step in to correct them. And again, as British journalist, Carole Cadwallader, said in a Ted Talk: "…this entire referendum took place in darkness, because it took place on Facebook. And what happens on Facebook stays on Facebook, because only you see your news feed, and then it vanishes, so it's impossible to research anything. So we have no idea who saw what ads, what impact they had, what data was used to target these people. Or even who placed the ads, how much money was spent, even what nationality they were."

Third, the algorithms. They can play a big role in promoting falsehoods. Roger McNamee, who was an early investor and advisor to Facebook, said this about the 2016 US Elections: "Facebook’s algorithms have played a huge role in this election cycle by limiting each member’s news feed to 'things they like', which effectively prevents people from seeing posts that contradict their preconceptions. Trolls on both sides have exploited this bug to spread untruths and inflame emotions."

Platforms’ algorithms have boosted conspiracy theories, false claims to the top of search results, recommendation lists and news feed rankings. By giving them prominence, the falsehoods are made to seem more credible. Ranking does matter. One study found that manipulating search engine results to favour one candidate over another can alter voter preferences by 20%.

This spread of falsehoods is also aided by service providers in a growing commercial disinformation industry. Cybersecurity firm TrendMicro gave evidence before the Select Committee of the going rates for tools and services. I have put it in your documents. [Please refer to Annex 3.] It spread falsehoods. One million Instagram "likes" – just pay US$18, you can get one million "likes". One hundred Twitter followers, likes, or re-tweets, 34 cents. One hundred YouTube subscribers, 66 cents. Make a falsehood appear on a YouTube main page for two minutes, US$621. Make false content appear on legitimate news sites as if it is real, costs a bit more, US$20,000. To use online propaganda to instigate a street protest in the US, US$200,000, but you can actually get a street protest. There are also "hired guns" who spread disinformation for their clients in return for a reward, a market opportunity.

In Latin America, you have Andres Sepulveda convicted of rigging elections throughout Latin America in return for payment from clients. He used the cyberspace to spread falsehoods on key domestic issues and policies. Indonesian authorities uncovered one such syndicate, Saracen Cyber Team. Just one. There are so many of them.

Next, who is misled by these falsehoods? In a 2018 survey by Ipsos, a global independent market research agency, on Singaporeans, 91% of Singaporean respondents could not correctly identify one or more of five false headlines presented to them. Falsehoods work. They work in a complex manner. They take advantage of the attention spans, mental shortcuts, cognitive biases of reasonable, rational people. So, every day, all of us decide which issues to give our time and attention. A lot of the information we come across, we make quick judgements, based on who it comes from, how aligned it is with what we know and believe. These are the human loopholes that falsehoods seek to exploit.

People’s socio-political identities can play a key role in why people believe and share falsehoods. It is consistent with psychological studies on confirmation bias. So, Members can see from all this that in the Digital Age, the conditions supporting a Shared Reality have fundamentally changed. From a main artery sustained by mainstream newspapers, public discourse is fragmenting into millions of social media groups and conversations. Mainstream media was likened by law professor Dr Thio Li-Ann to a public street. On this street, she said, "you might encounter not only friends, but a … variety of people engaged in a wide array of activities."

In this system, people are made to see and show civility towards viewpoints they may not otherwise like and choose to see. But social media does the opposite. It is designed to connect people with whom they want to be connected with, usually people they know or like, or who are like-minded. They allow people to cut themselves off from the views and information they do not like.

The Select Committee heard testimony from Dr Simon Hegelich and Mr Morteza Shahrezaye, the political data scientists. Dr Hegelich had advised German Chancellor Angela Merkel. They described the use of social media for public discourse as "an enormous misfit in design". On social media, communication is guided by private affinity and emotions. It is designed to be convenient. A world of likes, emoticons, memes and captions. But, as Dr Hegelich put it: "…political discourse should not be convenient. In democracies, politics should be the result of debates, which are often arduous, because a compromise between legitimate interests has to be found."

In the aftermath of the 2016 US Election, a group of academics from America’s leading universities gathered to reflect on the problem of fake news. One of their key conclusions was this, and I quote: "Current social media systems provide a fertile ground for the spread of misinformation that is particularly dangerous for political debate in a democratic society." For those who favour politics of polarisation, politics of misinformation over one of cohesion and reason, these conditions are ideal.

So, let us move on. Can we expect self-regulation by tech companies? Tech companies play a key role in all of this. They provide the platforms from which falsehoods and other content are spread. They have held themselves out as making the world a better place. Can they be relied upon to self-regulate? The evidence over the years shows clearly, no.

In Sri Lanka, Facebook’s years of inaction to hate speech is well-documented. Last year, rumours led to ethnic violence, mosques were burnt, people were attacked, a State of Emergency was declared. Facebook users lodged thousands of complaints over hate speech and Facebook did nothing. One post, and I quote, only one of many such posts: "Kill all Muslims, don’t even let an infant of the dogs escape." In the midst of on-going violence, what was Facebook’s response to this? "We have looked over the post. It doesn’t go against one of our specific community standards."

The Sri Lankan Communications Minister highlighted a tweet from this user as well. Last year, in London, Senior Minister of State Edwin Tong questioned the Facebook’s representative on this post. I would like to show the video of that questioning. With your leave, Sir. [A video was shown to hon Members.]

Can we rely on self-regulation? Two weeks ago, in Colombo, Sri Lanka, terrorist bombings, more than 200 people killed. The terrorist leader had made videos to gain thousands of followers, called for attacks against other religious groups, led to several previous attacks. But YouTube and Facebook removed the videos only after the Easter bombings. You see the slides of the photos of the victims. So, what do we say to them? "Sorry, we made a mistake"? As Mr Allan said to Edwin Tong, "Sorry, we made a mistake"?

Big Tech have continuously refused take down hate speech, ban groups promoting hate speech. As one American journalist put it: "Facebook’s world, Facebook’s rules." The fact is that the more users, more content on their platforms, the more user attention they can sell to advertisers, and the more they can profit. They are profit-making entities. Ferocious. Will they voluntarily change their business models? They have also created a permissive online environment for hate. You saw what happened in New Zealand. Something happened similar just a week ago. Copycat shooting in California in a synagogue where the whole thing was live streamed. Attacks in New Zealand were live streamed.

The Chair of the UK Digital, Culture, Media and Sport Committee, Damian Collins said, it was a "terror attack designed for social media", demonstrated why there had to be statutory regulation for social networks. Only now, Facebook has banned white nationalism from its platforms.

In the 2016 US Elections, 470 Facebook accounts used by foreign actors reached an estimated 126 million Americans. Facebook was questioned by the US Senate on whether they had plans to ban the purchase of ads using payments made in foreign currency. Facebook could not give a straight answer. Last year, I asked the same question during the Select Committee hearings. And, again, no straight answer.

Twitter, two years after the US Election, most of the accounts linked to the foreign disinformation campaign were found to be still active. The recent Indonesian General Election platforms helped disinformation spread, take root faster.

In November 2016, Mr Zuckerberg said, "It is a pretty crazy idea" to think that fake news on Facebook had influenced the 2016 US Election. Ten months later when there was clear evidence of serious foreign interference, he apologised for having being dismissive.

In November 2017, Facebook was questioned by the US Senate. They promised to do better but said they would prefer to self-regulate. The recent release of internal EU Commission documents showed that over the last four years, Facebook has pushed back against regulation; consistently said regulation will stifle innovation.

To our Select Committee in March last year, they took the same line. "Prescriptive legislation" would make it harder for Facebook to find the right solutions. So, what right solutions have they found?

In March this year, one day before we tabled this Bill in Parliament, Mr Zuckerberg said regulation against harmful content is needed, the problem too big for them to handle. That is what we are now doing. Facebook made this concession because it had no choice.

In February 2018, a UK parliamentary committee questioned Facebook on whether it had provided user data to Cambridge Analytica, which had used data mining to help political candidates and groups target potential voters. Mr Simon Milner, Facebook’s representative, said to the UK parliament Facebook had not provided such user data to Cambridge Analytica, did not disclose that the user data had been taken by Cambridge Analytica, and he knew of it at the time. I questioned Mr Milner about this during the Select Committee hearing. He then conceded he should have given a more frank answer to the UK committee. Agreed that a reasonable person could take the view that he had not been full and frank in his answers, and had misled the UK committee.

New Zealand’s Prime Minister Ms Jacinda Ardern, has said: "We cannot simply sit back, accept that these platforms just exist and what is said on them is not the responsibility of the place where they are published. It cannot be a case of all profit and no responsibility."

I also read Nominated Member Prof Lim Sun Sun's opinion editorial (op ed) on this point. Her view is that government, big tech and consumers should come together and work in a tripartite fashion rather than relying on legislation alone. Fair point. I have said on this subject as well as other subjects. Legislation provides a framework. By itself, legislation cannot achieve all the needed objectives. A lot of cooperation is needed. Tech companies are not our enemies. They are our partners. We want them to succeed. We want them to do well in Singapore and as well. And we must work with them to solve problems.

But at the same time we need to be quite clear about the dynamics. They are focused on profit. And there is little incentive for them to do things which will affect their profits. In Singapore, they must know, we must make sure they know there is public interest. They can make profit but not at the expense of Singaporeans. And POHA and POFMA seek to provide that framework. And they know Singapore Government cannot be bought. We do not take money from their lobbyists and we mean what we say. They can do business with us honourably. Singapore provides a proper rule of law framework for everyone but they must also be responsible.

Last year's Select Committee proceedings and their dealings with us would have left them in no doubt that they can push back when necessary. Throughout, in good faith, we have had constructive discussions with them. But let me give Members a recent example.

On 4 May, last Saturday, the Managing Director of AIC Jeff Paine published an op ed. AIC is a lobby group for tech companies. And his op ed was full of basic errors. One of the points he made. He referred to clause 61, the exemption clause, and said that this was troubling. This clause gives the power to exempt and really will help tech companies. Because where they cannot comply, they can show reasonable reasons why they cannot comply, exemptions can be given. Since they said this and we have been discussing in general how we can help tech companies when they cannot comply, I asked my Ministry to tell the tech companies, that was on Monday because the op ed came out on Saturday, I told my Ministry officials to tell the tech companies, "You do not think this clause is necessary, right? Okay, I am quite happy to remove. I am speaking in Parliament on Tuesday and if I do not get any response before I speak, I will assume that tech companies do not want this clause". Because how do I tell the Members of this Parliament that this is primarily to help tech companies when they themselves say they do not want it.

What was the response? It was quite swift. Last night, AIC issued a statement. They said they appreciated Senior Minister Edwin Tong's clarification on the exemption clause and they appreciated the Government's efforts to make POFMA technically feasible for tech companies. That is what they need to see from us.

These are not children's games. This is serious business. Tech companies will say many things to try and advocate their position. We have to show them we are fair but also firm; and then they will deal with us in a business-like way. There cannot be and I do not think there is any serious dispute that legislation is necessary. The question is what sort of legislation.

This Bill is based on the Select Committee report which explains what outcomes should be achieved. The Bill seeks to achieve those outcomes. We think the setting out in the Bill will best achieve the outcomes the Select Committee had identified. Senior Minister of State Edwin Tong will go through the specific provisions. I will now deal with some of the criticisms of this Bill.

Ms Han Yong May had a good commentary in Lianhe Zaobao some weeks back. She said: (a) there is nothing wrong with questioning whether the powers under POFMA will be abused; (b) but people should first read the Bill in detail. This is part of responsible speech and if they did, they would understand that certain things people have been saying about it are untrue; and (c) at the same time, she had a piece of advice for the Government. Please understand that some people may be genuinely concerned and do not see shadows and treat them as being purely antagonistic. I think fair points.

The following are some of the points that have been made about the Bill: (a) it gives too much power to the Government; (b) it will have a "chilling effect" on free speech — that phrase again; (c) definition of fact and opinion is not clear; (d) definition of public interest is too wide; (e) difficult to challenge the Minister’s decision.

Let me deal with each.

I explained to Members earlier. The Bill is narrower than the current law in terms of the power this Bill gives to a Minister. If a Minister declares that an article contains falsehoods and asks for a Correction Order to be carried, or in some cases, take down, it is open to challenge in Court. If Minister is wrong, he gets overruled.

And this inaccurate claim is then sought to be backed up by a series of other inaccurate claims.

First, it creates a new crime; spreading of falsehoods is a new crime. That is quite inaccurate. It already exists under section 45 of the Telecoms Act (TA), now under the MOA, and applies to the Internet. TA criminalises falsehoods; the Bill covers the same ground, falsehoods. But in terms of intent, it requires the person to have known or had reason to know not only that it was false, but also that spreading it was likely to prejudice public interest. So, in fact, there are two requirements for knowledge. And both requirements have to be satisfied, even though it is based on having reason to know. Whereas under the TA, it is just sending out of falsehoods.

Second, it is often said that recourse to judicial review is not available under the Bill. Short answer: wrong.

Third, a judge cannot examine the proportionality of a direction. Wrong.

Fourth, power in clause 61 of the Bill to exempt certain persons from the Act can be used for some unstated sinister purpose. This exemption power is a common provision in many pieces of legislation. It is found in the Executive Condominium Housing Scheme Act, Postal Services Act, Active Mobility Act, Dangerous Fireworks Act and a wide variety of laws. I explained, for example, when it is not possible to comply with the Act, due to reasons such as technical impossibility, then we could rely on this.

Next, some academics have said that: (a) truth is relative; (b) impossible to state what is a "fact" as opposed to what is "false and misleading"; and (c) this will affect academic work. Minister Ong Ye Kung will respond on matters specific to academia. I will just say this.

Their concerns appear to be expressed without an understanding of the existing legal position. How this Bill fits in. And I am not sure how many of them are aware of the BA, the TA, the MOA and other legislation.

Just to give you a couple of examples. A law lecturer signed the letter. Did they know know the current legal position? Another signatory, in a recent interview with Civicus, an international NGO, made several completely erroneous statements. He said the Bill defines a "falsehood" tautologically, such that any Government Minister can define a statement as false based on "public interest". These are two different and conjunctive requirements: you have to show it is a false statement of fact, and you have to show it affects public interest. He, unfortunately, seems to conflates the two and says that if it is contradicts to public interest then we can declare it as false. Of course, if you believe that then you will get very worried. I read the interview more than once to try and see that is what he really meant. But that is what the words say. It is very odd. A fairly fundamental error. If that is what the Bill states, it will be quite illogical – "it is false because public interest is involved".

If a Singaporean academic can be mistaken on such a basic point or a set of points, then it is understandable that foreign academics would have signed without an understanding of appreciation of the true position. As I have said, laws that target false statements or facts have been around for a long time.

Meanwhile, in the last 15 years, for the Times Higher Education World University Rankings, NUS went from 18 to eight and NTU went from 50 to 12. There is no legal or logical basis for the concerns that this Bill will stifle academic research specifically. As I said, Minister Ong Ye Kung will respond later on some points.

Then, it has been said: why not define "fact" and state that "opinion" is not covered. We have considered this carefully and decided better not. There is a body of case law on what is "fact" and what is not fact. It is better to rely on existing case law. When there is a dispute, the matter can be dealt with by the Court. So, how else do you decide?

Next, chilling effect. I think this must be one of the most overused phrases. Free speech should not be affected by this Bill. We are talking here about falsehoods, we are talking about bots, we are talking about trolls, we are talking about fake accounts and so on.

Prof Thio Li Ann gave evidence during the Select Committee hearings on the ambit of free speech and whether falsehoods were free speech. She made the following points: not all forms of speech are worthy of equal protection. For example, if you falsely cry "fire" in a crowded theatre, that is not protected as valuable speech; UK House of Lords' judgement: "There is no human right to disseminate information that is not true. No public interest is served by communicating misinformation. The working of a democratic society depends on the members of that society being informed and not misinformed. Misleading people, purveying as facts statements which are not true is destructive of the democratic society". Where speech does not serve the justifications for free speech, by harming the search for truth or by preventing citizens from becoming informed of issues – that does not warrant protection.

Next, the definition of "public interest" is too wide. It is important to remember that the Bill does not cover statements just because they are against the public interest. Those statements must be false statements of fact in the first place. Some have said that the definition of what is in the public interest is too wide, and that it should not include clause 4(f), which relates to the diminution of public confidence in the functions of Government institutions. I have explained earlier in some detail, maybe too considerable a detail, on how slow-drip, online falsehoods seek to break down trust by attacking institutions.

It is important to protect institutions from such falsehoods. And if you look at the point I made many times, you look at the current definition and you look at the existing definition, you decide for yourself which is wider. And then it is said it is difficult to challenge a Minister's decision. I have previously said, the process that will be in place, it will be fast, it will be simplified to allow individuals to appear and present arguments on whether it is true or false. The detailed procedure will be in a subsidiary legislation and that is usual. But I will set out an overview of the process.

First, an aggrieved person must apply to the Minister to cancel a Direction. This is consistent with the usual position of exhausting administrative remedies before resorting to the Courts. We will provide a standard form online for aggrieved persons to use. The form must be sent to an email address which will be set out in the Direction and the relevant Minister must make a decision no later than two days after the form is received, excluding non-working days.

The appeal to court will be similarly quick. The appeal will have to be filed no later than 14 days later. This is up to the applicant. He can file the very next day if he wants, after the Minister decides on the application to cancel a Direction. Simple standard forms will be provided for the appeal documents that an appellant can fill out and file in the Courts. The Courts will be asked to fix the hearing within six days of an application being filed. The appellant must attend before the Duty Registrar to request for an expedited hearing, again in the manner prescribed by the rules of the Court.

The documents will need to be served on the Minister no later than the next day. Again, an email address will be provided for the appellant to serve the documents on the Minister, to make it easy for him. The Minister must then file his or her reply in Court no later than three days after the documents are served as prescribed. As stated earlier, meanwhile, the Court would have already fixed the hearing no later than six days after the date on which the Court first received the application.

To summarise, a person aggrieved by a Direction will have the opportunity to have his or her case heard in the High Court as early as nine days after he initiates a challenge by writing to the Minister.

So, this includes the time the Minister responds, the time for filing, the time for him to file and the time for the Minister to respond and the hearing, that is, beginning to end. If he moves very fast, it can be nine days. They will have to be working days. The Court will continue, of course, to have a general discretion to extend timelines where there is good reason to do so. How long the hearing takes and how long the Court takes to decide are matters for the Courts. Parliament and the Executive cannot intervene in that. We can say when it should be heard; we cannot say when it should be decided. Court fees will be kept very low for individuals. No hearing fees will be charged for the first three days. Further days of hearing will be charged at the usual rate. But even then, the Court will have power to waive fees. But this should not be taken as a licence to abuse the process. Courts will still have the power to deal with parties in the usual manner, including how they conduct themselves.

Assoc Prof Eugene Tan of SMU suggested we use illustrations during the Second Reading to explain the difference between fact and opinion. Good suggestion. So, I asked for some illustrations from groups which we met – CAPE and students from NUS and academics. The list is enclosed and circulated to all Members and will form part of the record. [Please refer to Annex 4.] I will refer to some additional examples.

A Professor of Law states that the death penalty in Singapore does not deter crime, as shown by several studies. These studies present real data. That is a conclusion drawn from the studies. It is a statement of opinion not covered by the Bill. But if that professor refers to a non-existent study, or non-existent data, then it is a false statement of fact.

This example is drawn from a recent incident and was one of the illustrations submitted. A person, say, "A" states that the Government is showing "double standards" by having two different permit regimes to govern public activist events on political issues on the one hand, and a public meet-up by an influencer on the other. It is true that two different permit regimes apply to the two different kinds of events. Therefore, "A's" statement is an opinion and is not covered by the Bill.

However, if "A" says or implies that the same permit rules govern both kinds of events and double standards were applied for the granting of permits, essentially suggesting bias – based on the point that the same legislation and rules apply for both. That is a false statement of fact. It can be covered by the Bill. The remedy would usually be a clarification. Forms of clarification will be set out in subsidiary legislation; basically a link to where the clarification can be found with the statement that the article contains inaccuracies or is false, and reference given to the link to the true facts. And people can read both documents and decide for themselves. It actually encourages a more honest and more open discussion.

Another illustration: an economist states that the home-ownership rate in Singapore is only 9%, and explains that the definition of home ownership used excludes any lease, regardless of whether the lease is nine years or 99 years. That is a conclusion based on a certain methodology. He says that and he sets it out. It is not covered by the Bill. The Government can, of course, disagree with that view. If, however, the economist says that the number of persons who own freehold property in Singapore is X, when it is actually Y, then that is a false statement of fact.

Another illustration: "B" publishes an online article stating that the social welfare system in Singapore has gaps. That conclusion is drawn from interviews with Government officials and social workers. The Government can disagree with the conclusion, but it does not come within the Bill. If, however, B's article says that welfare assistance had been denied to a needy elderly person when assistance had in fact been given, then it is a false statement of fact. So, Members can see that it will be clear when you look at examples. If public interest is affected, then it can come under the Bill. In all of these, it is not enough to show it is false. It must also be affecting public interest. This is a situation where, again, the original article can carry a clarification.

I will use another recent incident to illustrate. In the case relating to Mr Nicholas Lim, there were questions as to why he was not charged. The Police investigated the case, gave a conditional warning. Some made allegations that Mr Lim was not prosecuted because his parents were influential. That was false. His father was a driver in the public transport sector, mother a housewife.

So, the illustrations, once you look at them, I think they are quite clear. The Courts have long dealt with this. It really ought not to be an issue. And we have had this legislation in the books forever.

Next, let me turn to the Notice of Amendment put by the Nominated Members of Parliament. I think it is useful for me to set out our views on the NOA. As I said elsewhere, there is substantial agreement between the Nominated Members and us on the major points of the Bill. The differences we have on the Directions' regime are mainly on matters of process. They have made four proposals.

First, the three Nominated Members have proposed that the Directions provide more specific reasons on falsity and public interest. I agree with the principle underpinning the first part of this proposal. The Government, when it says something is false, must obviously set out the reasons why it is false. And we intend to set this out in subsidiary legislation. When the Directions are given, there must be a statement as to why it is false. And this was intended all along. The legislation is specifically designed for this purpose. Clause 13 allows for subsidiary legislation to be made to set out the requirements for Directions.

It is a question of operationalisation. When a Minister issues a Direction, what kind of information should be contained in that Direction? Why something is false, as I have explained, should be explained in the Direction. What should be the level of detail? I think it is difficult to set out upfront. It must be left in each case what extent of detail should be specified. It is really more appropriate to use subsidiary legislation to set out the requirements. And subsidiary legislation is law and it is public, transparent and open to scrutiny. Separately, on public interest, again, I am not comfortable requiring a setting out of detailed reasons on how a particular course of action serves public interest. The level of detail depends on each case.

It is difficult for Parliament today to envisage what are the types of cases. How are you even going to start setting out what level of detail ought to be set out? You do not even know what the case is about. But you must set out enough detail and, if not, you take the risk of a Court challenge and then you go and explain to the judge. If there is a challenge, the Government will have to show public interest and we will have to explain in Court why the information provided is adequate and depends on whether the Court accepts the explanation. That is the best way of dealing with these things. You cannot go upfront and try to legislate every detail.

Second proposal is to set out in the primary legislation, that appeals to the Minister must be adjudicated without delay. I agree with the policy. It is just a difference on how it should be implemented. Timelines are usually set out in the subsidiary legislation and I have set out what the timelines are that we are going to put in place. This will be in law, by way of subsidiary legislation. And we are going further than what the three Nominated Members have proposed. Their amendment proposes the "Minister shall do everything reasonable to ensure that appeals to the Minister are adjudicated without delay.".... "High Court hearings shall commence as soon as practicable, and that costs to the appellant are minimised”.

Assuming you put these broad principles in legislation without any specifics, what is "as soon as practicable" and what is "without delay"? Court hearings today can take anything within 12 to 18 months, sometimes longer. So, if we had it in six months, is that "as soon as practicable"? Obviously so. So, one should avoid this sort of imprecise language in legislation because Minister shall respond "as soon as practicable" – is that 12 days, 14 days or 30 days? How are you going to challenge? How are you going to say, "Well it takes six months for the hearing to take place." The short answer to that will be, "Well, you know, it is "as soon as practicable" compared to the fact that it takes 18 months now." You really do not want to go into that territory. What you want are precise timelines, the kind of timelines I have given. From the time you send the Notice to the Minister to the time you are in Court, very precise. If you get your act together, nine days. Much clearer.

So, I am arguing against the Government being given more time. I do not think it is good enough simply to say do everything reasonable to decide quickly. You must mandate that the Minister decides within a certain time period and it is not enough for the High Court hearings to commence as soon as practicable. We should set out when the hearings should start.

The third proposal made by the Nominated Members is the proposal to insert key principles into the Act. The three Nominated Members want to insert certain principles. I appreciate the intent underpinning the proposal. But, frankly, without any disrespect, some of the proposals are unworkable and some others are unsuitable for primary legislation. I can understand that they are not lawyers so let me just take them through.

If you look at para (a), it covers the importance of critical thinking to a democracy. The principle is important. It is acknowledged by the Select Committee in recommendations 1 and 2. And it is one of the tools that supports the development of a well-informed public. But you need legislation at the same time and you have got to have legislation work with non-legislative measures. But in the kind of situations we are talking about, there is no question whatsoever that legislation is essential and this Bill is about legislation.

Separately, the Government has committed to a whole series of non-legislative matters; and Parliament can question the different Ministries ranging from education to public education to media literacy, there is a a variety of things. But when there is public interest involved and there is falsehood, and the Nominated Members accept the Bill is necessary, then there can be no argument and the Bill has to be used and the powers under the Act have got to be used. It is necessary then to push out the facts to the public by way of correction to inoculate them against the falsehood.

The evidence is set out in the Select Committee Report – if you look at paras 390 to 396 and 425. You do not get much by adding this to the legislation. You have got to be very clear. The legislation is for this purpose.

Then, there is a group of, I think, proposals which target, a sort of negative list about when statements are not false and when statements are not statements of fact. I understand the intent which is to clarify matters. But if you look at your proposals (b) and (c), they essentially relate to the process of establishing truth and falsehood. As I have said, "Have a care". There is an entire body of law in this area. The law of evidence is there to deal with the question of how facts should be established. The question of how truth or falsity should be established and it should draw from the body of principles that form part of case law. It would not serve the rigour of this process to point out some but not the others. Then, you go and start putting this thing in, you run the risk of trying to take out or deal with some parts of an existing body of law. The Courts will deal with it where necessary.

Similarly, on principle (g), I have already explained why it is preferable to rely on the existing case law.

The third group of statements proposed in (d), (e) and (f) and they essentially relate to proportionality. Proportionality is already incorporated into the requirements of under the Bill and it is stronger than your proposal because when it is incorporated in the Bill, it is part of the legal requirement. Yours are only statements of principles. This is already in the law. So, the legal test already requires that the power can only be exercised when it is necessary or expedient in the public interest to do so. Government has gone one step further, perhaps more than one step further, than what you have proposed.

Finally, principle (d). The Bill only deals with situations where there is a falsehood which affects public interest. So, it would not make sense to suggest that we insist upon non-legislative measures when public interest is at stake because there is a need to respond quickly to counteract the damage. So, does that mean by putting this in that you do not ask for corrections?

And I caution you, it can lead to endless litigation. Let us say the Minister seeks to act under the Bill and there is a principle here; then it can be argued, "Well, you know, in this case you shouldn't act because it's not necessary. You can take non-legislative measures." Someone argues that. It may fail but that can take months. It can create a fair bit of confusion as well. So, we should just be very clear what the Bill provides, what is the framework for the Bill and try not to over-egg the pudding. We run more risks that way. We should have a robust debate here, explain what the Bill is all about, have it as a matter of record in the Hansard and that is used by the Courts as a matter of interpretation.

So, the Nominated Members' concern about enshrining principles in a manner which would bind future governments, I think really is best served by the debate and by the Courts serving as an oversight mechanism and interpreting the provisions in future. But, you know, a future government can always amend. So, that is a separate point.

Finally, there is a proposal for an Independent Council – the fourth amendment. I understand the intent again. It is a good attempt to ensure that the conversation on this important issue continues both on the nature of the problem and how the legislation is implemented. The Government agrees that these conversations should continue.

The question is, is establishing another council the best way? Would it lead to unnecessary bureaucratic bloat? All of the functions of this council can be achieved under the current structure of the Executive and Parliament. On accountability, Ministers, the Executive, are responsible to Members of this House, to Parliament. Every time there is an exercise of power which seems to you to be not clear, you ask questions. You just look at the questions that have been filed since you became Nominated Members, every major issue, anything that happens, from NUS to other areas, you do not need to wait for one year. It is raised immediately and answers are given.

Remember what I said about authenticity. The Singapore public must know this institution works and if there is a question, it will be answered. There has got to be the trust as well and where the trust is not merited, that has got to be exposed as well. Parliament is the representative body. It is meant to work in a democracy in this way where Members of Parliament hold the Ministers accountable and check and ask questions. You do not keep creating bodies and then who checks on them. So, it is really more effective to rely on the strength of our current institutions – direct accountability. And together with accountability with the Courts. So, you have Parliament, you have the Courts. What more do you want? Any time there is a question, it can be challenged in Courts, it can be challenged in Parliament. If you keep getting it wrong, the electorate will have something to say about it.

So, I think I have covered most of the major arguments. I have set out the nature of the philosophical reasons for the Bill at some length –perhaps too considerable a length, for which I apologise. With that, Sir, I beg to move.

*Question proposed.*

[Deputy Speaker (Mr Charles Chong) in the Chair]

The speed with which online falsehoods spread and threaten public interest cannot be addressed effectively using the usual judicial process and require timely executive action. We also cannot rely solely on non-legislative measures, such as self-regulation of online platforms or only focusing on promoting media literacy. Countries that rely on such approaches have not found much success in tackling fake news.

Battling fake news is similar to combating terrorism or fighting a war on drugs. There is no hard and fast answer. However, neither should a blanket policy be applied. Formulating a solution will require vigorous debate and consideration before executing a final calibrated response, much like a surgeon using a scalpel to remove necrotised tissue and not harming healthy cells.

Minister Shanmugan shared that the current existing laws give the Government more powers and are broader. In view of the recent public concerns, should we not then seriously review this, rather than using existing laws as gold standards, because those are big hammers and I believe they are intended for a different purpose?

Currently, I am aware of the following countries that have enacted laws against falsehood. For example, Germany’s Network Enforcement Act or the French anti-fake news law against the manipulation of information. In both instances, a "notice and take down" order is issued against purveyors of fake news.

The Minister has suggested that the Bill put forth offers a more calibrated approach wherein correction orders are possible. This allows for contents with correction notice to be still available for debates and, therefore, more informed decision making and this is, indeed, progressive.

However, one feature to note in our current Bill is that it applies to any person or organisation deemed to be spreading fake news. The aforementioned German and French legislation limits the scope of the law to only tele-media service providers operating Internet platforms. Large tele-media companies have the ability to counterbalance and respond to challenges from the Ministry. It is quite different for individuals or smaller organisations and, hence, the concerns about the impact on general freedom of speech.

I would also like to highlight some concerns regarding the definitions listed in the Bill. The definition of what constitutes "statement of fact, public interest and diminution of public confidence" is very broad. Sir, I suggest that these definitions and terminologies be clarified and explicitly defined for the layman in the principles of the Act in the Bill, as what has been practised for POHA. This will help agencies effectively implement the Act and give the general public a general grasp of what it means. The law cannot be set up just to be understood by the Court and citizens do not understand at all.

The consequences of these overly broad definitions can lead to chilling effects on freedom of speech and expression for our country. Our citizens will need to think twice if a view is a fact or an opinion before expressing it and be able to correctly interpret a view even if information is not readily available.

This issue is compounded by the fact that Singapore does not have a Freedom of Information Act, where citizens can request for data from the Government instead of having to do their own fact-checking and analysis. Can we remedy this by providing information that the public seeks? Would the Minister consider implementing both Bills together as a counterbalancing approach?

Citizens would also need to consider if the content they share or upload is targeting a person or the Government and take precautions to avoid legal consequences with laws, such as the Defamation Act or the Internal Security Act, amongst others. How would these Acts cooperate or conflict with each other, given their overlapping functions and how will they be handled?

Depending on how this law is implemented, there will be heightened concerns with legal ramifications. Apprehension and over-cautiousness will arise from expressing views and uploading content for fear of swift legal reprisal. Citizens will feel inhibited to raise or discuss sensitive topics related to the issues close to their hearts.

There is also concern of the legislation being used against citizens who criticise the Government or have views contrary to the ruling party. Law Minister Shanmugan has clarified in the Straits Times that opinions, criticisms, parody and satire will not be covered. However, these reassurances are not explicitly addressed in the Bill. Proactively including this language will go far to allay these concerns.

Sir, this brings me to the next point to confer Ministers with executive powers to issue correction orders and take-down notices. How do we ensure consistency in approach across Ministries to determine between fact and opinion? How do we see to it that there is no "cherry picking" when it comes to acting against specific information not in favour of the Government? If fake news originates from the Ministry itself, who then decides to have that Ministry take it down? This is the characteristic dilemma of "Who watches the watchers?" In a democratic society, the rule of law should apply to everyone.

I acknowledge Minister Shanmugan's assurance that members of the public have the recourse of appealing the Government's decision via the Court and he has also advised that the process will be relatively "fast and inexpensive".

That said, there is no mention in clause 17 of the Bill for an expeditious recourse. The Bill does not specify the prompt access to recourse in the event of a wrongly executed take-down of supposed fake news. I would, therefore, like to propose a stipulation in the primary legislation for the Government to respond in a timely fashion for information clarification and facilitation of the appeal process.

Although I understand a timeline for appeal will be stated in the subsidiary legislation, I strongly urge the need for an expeditious appeal process to be included in the primary legislation, for example, not more than how many days. This signals a stronger reassurance that prompt access to remedy is readily available and any future Government will have to go back to Parliament to have this amended.

The next point of contention is the lack of an independent oversight council for checks and balances. Though the final arbiter will be the High Court, what about cases that do not appear before the High Court? This is pertinent if actions will be taken against individuals or small groups of people compared to large tele-media companies, whose access to resources provides greater efficiency for recourse in Court.

This oversight council is in line with the Select Committee's Recommendation 10c: "Existing efforts should be reviewed, to consider whether they are adequate to achieve transparency, participation and communication and accountability". Having an independent oversight council to evaluate the effects of fake news will allow directives to be issued in an effective manner and ensure credibility in our governance.

The proposed council is not meant to be at odds with Ministries or the Courts in any way, but to allow for transparency, accountability and synergy. It will clearly add much-needed integrity to our entire political system.

Additionally, the case studies compiled from the council's annual report can be used for public media literacy education, and to establish consistency across ministries in refining future Bills. The Minister has said that the Parliament should be the final check and balance, and in this place, should we then not dissolve all the existing councils. I believe they are set up for a reason. The same rationale for our proposal.

As the first country in the world to enact a Bill with such broad strokes to curb the volatility of fake news, a steep learning curve is to be expected. I believe that we will get it right, striking a balance between curbing fake news and ensuring freedom of speech.

Central to the discussion is the element of trust in pursuit of truth. The majority of Singaporeans support the vision and bravery of this Bill and trust the sitting Government to be honest and just in its implementation. As implicit in the asking of trust from the people, trust should also be paid forward to the people, that we are able to discern fake news as a highly educated populace. The Government will imperil itself if the people perceive that this Bill takes too much liberty with their freedom of speech. I appeal to the House to seriously consider lifting the Party Whip and allow Members to vote according to their conscience for tomorrow's voting. This Bill will have a huge impact on the average Singaporean's freedom of expression and possibly our nation's progress.

Many civil groups harbour genuine concerns over this Bill, as evident by the 170 written representations from individuals and organisations submitted to the SC, and this includes numerous open letters and joint statements published in protest of this Bill from both local and international fronts.

Hence, I think it is our moral obligation to give constructive criticism and proposals to fine-tune this Bill in the best interests of Singapore. If left unrevised, this Bill may have long-term repercussions on freedom of expression and speech, impede the development of an engaged and intellectual public, and erode the trust in our Government. At the end of the day, it is our solemn duty to uphold the mandate by the people and justify their trust through uncompromising policy-making and rigorous debate. Undoubtedly, I certainly have strong reservations for the Bill in its current form.

While I do see positive elements within the Bill, Mr Speaker, this Bill in its current state will need further clarifications before it is ready to be passed. I would like to receive the response to my questions and an honest discussion following.