IMDA rolls out transparency guidelines for GenAI chatbots

The Infocomm Media Development Authority (IMDA) has introduced a new set of voluntary transparency guidelines for generative AI (GenAI) chatbots, aiming to help consumers better understand how AI-powered applications work while encouraging organisations to adopt clearer disclosure practices.

The guidelines, which IMDA said are among the first of their kind globally, seek to establish a consistent framework for how organisations communicate the capabilities, limitations and safety measures of consumer-facing GenAI chatbots. The move comes as AI chatbots become increasingly integrated into everyday services across sectors including banking, healthcare and travel.

At the centre of the guidelines is a "chatbot info card", a standardised disclosure that consolidates key information about an AI chatbot in one place. Similar to a nutrition or medicine label, the info card is intended to help users quickly understand what a chatbot can and cannot do, how reliable it is, how personal data is handled, and how concerns or issues can be reported.

Don't miss: Are AI chatbots building the next walled garden?

According to IMDA, the information should be relevant, written in plain language, easy to locate, and kept up to date whenever significant changes are made to a chatbot's capabilities, risks or safety policies.

The authority said the guidelines are designed to help consumers make more informed decisions when using AI applications, establish a more consistent approach to transparency across different chatbot services, and encourage organisations to clearly communicate their safety commitments.

The framework was developed following consultations with technology companies, enterprises operating consumer-facing chatbots, government agencies and members of the public to understand current industry practices and the type of information users consider most valuable.

According to IMDA, several organisations have indicated they intend to reference the guidelines over the next six to 12 months as they strengthen transparency practices for their public-facing chatbots. These include Google for its Gemini app, Meta, as well as Singapore-based organisations including DBS, OCBC, Singapore Airlines and Synapxe.

Meanwhile, sector regulators are also aligning with the broader push for responsible AI adoption. IMDA said its transparency guidelines complement existing AI governance frameworks issued by the Monetary Authority of Singapore and the Ministry of Health, both of which emphasise transparency as a key principle for AI deployment in financial services and healthcare respectively.

The public sector is also expected to adopt the framework, with agencies such as the National Library Board and Health Promotion Board planning to reference the guidelines for their own consumer-facing chatbots.

While the initial focus is on GenAI chatbots due to their widespread consumer use and growing concerns around issues such as data privacy, child safety and risks to vulnerable users, IMDA said the principles can also serve as a reference for other GenAI applications. It added that it will continue working with sector regulators and international partners to adapt the framework for other industries as AI adoption expands.

The latest guidelines build on Singapore's broader efforts to strengthen AI governance and establish international standards for responsible AI deployment. Earlier this year, Singapore proposed a new international standard for testing GenAI systems, a move that could shape how brands, agencies and technology providers deploy AI across marketing, content creation and customer engagement.

The proposed ISO/IEC 42119-8 standard focuses on benchmarking and red-teaming methodologies for GenAI systems, with the aim of standardising how organisations test AI models for trustworthiness, safety and reliability. The proposed framework is designed to establish a common approach to testing GenAI systems, improving the reproducibility and comparability of results across organisations while supporting safer and more trustworthy AI adoption by businesses and end-users.

Related articles: 
AI chatbots are getting 'aggressively' data hungry   
Brands struggle with AI disclosure as usage surges across marketing  
More companies miss revenue targets as AI and volatility reshape B2B growth