Singapore announces new AI safety initiatives

One of the initiatives is the Global AI Assurance Pilot.

Singapore has launched new artificial intelligence (AI) governance initiatives aimed at enhancing AI safety for both Singaporeans and global citizens, acknowledging the transboundary nature of AI products and services.

These initiatives were unveiled by Digital Development and Information Minister Josephine Teo at the AI Action Summit (AIAS) in Paris on 10-11 February.

The new initiatives include the Global AI Assurance Pilot, launched by the AI Verify Foundation and IMDA to establish global best practices for testing generative AI (GenAI) applications. It will bring together AI assurance vendors and companies using GenAI solutions.

Another key initiative is the Joint Testing Report with Japan, which was released in collaboration with Japan under the AI Safety Institute (AISI) Network. The report focuses on making Large Language Models (LLMs) safer in various linguistic environments, assessing the performance of AI safeguards in 10 languages and five harm categories, addressing gaps in non-English language models.

Lastly,  the Singapore AI Safety Red Teaming Challenge Evaluation Report evaluates LLMs for cultural bias in non-English languages and provides a consistent methodology for testing across diverse languages and cultures to address regional AI safety concerns.

Teo also endorsed the Leaders’ Statement on Inclusive and Sustainable AI for People and the Planet. She discussed AI risks and governance at the summit, highlighting Singapore’s approach to balancing AI innovation with necessary safeguards.

She also participated in closed-door meetings with international policymakers and industry leaders, furthering Singapore’s efforts to shape AI standards and ensure flexible governance that can adapt to technological advancements.