Guidance for Muslims who Turn To Generative AI for Religious Answers

Mr Saktiandi Supaat asked the Acting Minister-in-charge of Muslim Affairs whether MUIS will issue guidance to help Muslims who use generative AI and other online sources to provide religious answers, to verify such answers against authoritative sources and to know when and where to consult qualified asatizah.

Mr Speaker, generative artificial intelligence (GenAI) and online platforms can complement religious learning by making information and knowledge more accessible. However, they should not be relied upon as the primary source of religious instruction or guidance, particularly on complex religious matters that require scholarly interpretation and contextual understanding. In these situations, a qualified asatizah, or other trustworthy sources of knowledge, should be consulted.

The Islamic Religious Council of Singapore (Majlis Ugama Islam Singapura or MUIS) has been reinforcing these messages through Friday sermons, public talks and the MuslimSG platform. MUIS will continue efforts to deliver practical guidance through social media and our network of community partners to strengthen digital religious literacy.

Mr Saktiandi Supaat.

Thank you, Mr Speaker. I have two supplementary questions. I thank the Acting Minister for his reply. Someone seeking religious guidance may not have the knowledge to verify an AI-generated answer, even when the references are provided. So, my first supplementary question is, could MUIS strengthen its existing channels so that members of the public can submit such answers for checking, for example, by qualified asatizah and receive timely responses, particularly on matters affecting important decisions related to religious questions on family or personal decisions.

The second supplementary question is how would MUIS ensure that the guidance reaches Muslims or Malay/Muslims who obtain most of the religious information online, as you mentioned, including those who may not regularly attend mosque programmes or religious classes? Will there be, for example, practical examples that MUIS may do, going forward, to help them recognise when an AI-generated answer needs to be checked and which ones would need to be checked by a qualified asatizah?

Thank you, Mr Speaker. Those are certainly valid questions and something that is of concern, not just to the Muslim community, but I think the wider community, in terms of how one consumes content today, which is available worldwide. The influences come from global influences as well.

So, I think, the issue is really about how we make our not just Jemaah, but ummah also, in terms of the wider citizen pool, better understand the need for being more aware to be able to check information, whether religious content or otherwise. I think, ultimately, that is something that we need to work on as a country.

Two things: one, to the Member's question of can we make our platforms more accessible, the answer is certainly. There is certainly room for us to improve accessibility to information, accessibility to asatizah, for example. This is one area in which MUIS is looking into, in terms of upgrading the MuslimSG platform.

GenAI is something that we look quite closely at. So, we are educating and even upgrading many of our asatizah to be more familiar with GenAI and new digital tools. Some of them have already undergone training to better understand the effects or how to use AI better, because ultimately, they do have to explain.

Secondly, I think the wider issue is how do we get the Muslim community to better engage the platforms, both physical and online platforms. Because I think the way we look at religious education has to be a holistic approach, both in the physical realm and digital realm. The asatizahs today are accessible, you can reach them via the Asatizah Recognition Scheme website. But certainly through MUIS, reach out to us if there are any concerns or any religious guidance that you seek. This is one area in which, perhaps, in the future as we upgrade MuslimSG, we could make it more interactive and more accessible for the community.

I think, ultimately, that is one area in which we are very cognisant about, in trying to reach out to the young especially, who may be influenced, especially where you have the risk of radicalisation and even what we call "ajaran sesat" or false teachings.

Within the asatizah community, I think the ability to contextualise some of the complex issues today that many get or receive or are influenced by from overseas clerics or Muftis. I think those are important aspects, in which we need to be able to help the community also contextualise many of these teachings back into the Singapore context too.

So, there are many layers that need to be worked on and that we are working on today. But I assure the Member that this is one area that we are looking at. And the online committee that I announced yesterday is also tasked to look into some of these emerging risks that come from online sources and online harms. We are certainly looking at it, both from a present state, and as well as proactively, looking at emerging risks too.

Dr Wan Rizal.

Sir, I thank the Acting Minister for sharing approaches on how we want to connect with the community. But I feel that what we do is pretty defensive. Is there some way we could create a dedicated unit, somewhere in MUIS, to comb the Internet and search for such religious misinformation in advance, before it gains traction to the community?

I thank the Member for his question. First, I think in an ideal state, you would have unlimited resources to do that. Because the Internet is very vast. It is very difficult to pick up every single site or every single source that comes along. Overall, what is more important is that we pick out themes, themes that resonate with people and sometimes, these are themes that resonate, but at the same time, may be false or not contextualised to Singapore's context. And we would have to explain some of these teachings.

For example, there were incidents a few years ago, where there were concerns of whether you should be able to wish others "Merry Christmas", can you wish someone that? So, certain themes which goes against the spirit of inclusivity, multi-racialism, our multi-religious context, have to be addressed.

So, we should proactively work on some of these themes, emerging risks that we see, and especially in terms of being able to keep people aware of online harm. Because they may not necessarily come from sites. GenAI does not come from sites as well; it comes from algorithms that are pulled together from different sites, so it is very difficult to police algorithms.

Ultimately, it is also about the same risk as we spoke about yesterday in this House, where you could have a self-styled influencer giving information too. So, it is not about guarding where the information comes from, but it is really also about inoculating our people, in terms of how to recognise the risk, helping parents to recognise the risk with what is going on with their kids. The broader picture is how do we help Singaporeans, especially Singaporean Muslims, contextualise many of these teachings, from outside, into our own multiracial, multi-religious context too.

So, there are different layers at which we need to work on. And certainly, we will work with the community as we move forward.