peopleimages.com - stock.adobe.c
Singapore’s AI Verify Foundation and Infocomm Media Development Authority (IMDA) have developed a draft model governance framework for generative AI (GenAI), expanding on IMDA’s earlier framework for traditional artificial intelligence (AI) systems.
The framework integrates ideas from an earlier discussion paper by IMDA and Aicadium, an AI technology company founded by Singapore state investment firm Temasek, that outlined the guardrails needed to address the risks of GenAI, among other areas, and builds on earlier efforts to provide an initial catalogue and guidance on evaluating the safety of GenAI models.
It also draws on practical insights from ongoing tests conducted in the GenAI evaluation sandbox that IMDA and AI Verify Foundation launched in June 2023 to let global AI players test use cases and evaluate trusted AI products.
Some of the proposals in the draft framework include tapping privacy-enhancing technologies to allow data to be used in the development of AI models while protecting data privacy; implementing processes for reporting and remediating AI safety incidents; using digital forensics tools to identify malicious code in GenAI models; and applying digital watermarks to provide transparency on where and how content is generated.
IMDA and AI Verify Foundation, whose members include Google, Microsoft, IBM, Red Hat and Salesforce, noted that the core elements behind the proposals are based on principles that decisions made by AI should be explainable, transparent and fair, providing practical suggestions for model developers and policymakers.
Noting that AI governance remains a nascent area, they stressed the need to build international consensus as demonstrated by the mapping and interoperability of national AI governance frameworks between Singapore and the US through the IMDA and the US National Institute of Science and Technology (NIST) crosswalk.
“The proposed Model Governance Framework for generative AI takes this one step further by covering the latest developments in generative AI,” they said. “In turn, this will inform Singapore’s next steps, as we adopt a practical approach to maximise both trust and innovation. This framework will evolve as techniques and technologies develop.”
Comments on the draft framework should be sent to [email protected]. All submissions should reach IMDA by 15 March 2024.
In December 2023, Singapore updated its national AI strategy to focus on driving more widespread adoption of AI; spurring people and businesses to operate with the ambition to be world-leading in AI; and working with stakeholders in and outside Singapore to exchange ideas and administer AI-enabled solutions at scale.
Read more about AI in APAC
- Indian organisations are speeding up deployments of AI across multiple sectors, but legacy systems, siloed data and a shortage of AI-specific talent will stand in the way of greater adoption.
- AI can be transformative in government, but its implementation has to be done in a way that drives positive outcomes while mitigating its downsides, according to the head of Australia’s Digital Transformation Agency.
- The multidisciplinary nature of AI offers career opportunities not only in builder roles like engineering and data science, but also in AI ethics and applied AI.
- Alibaba’s SeaLLMs are built to address the linguistic diversity and nuances in Southeast Asia, enabling businesses to deploy localised chatbots and translation applications.