What benchmarks and tools will be developed under this AI governance partnership?

Question

Accepted Answer

Benchmarks and Tools for AI Safety and Security Evaluation Will Be Developed Under the Partnership

The partnership between Singapore's Infocomm Media Development Authority (IMDA) and Microsoft Corporation will involve the joint development of benchmarks, tools, and other evaluation methods. This initiative aims to advance the field of evaluation sciences specifically within the context of AI safety and security. The collaboration seeks to foster innovation while ensuring that technology is developed and deployed in a safe, secure, and trustworthy manner.
Advancing Evaluation Sciences in AI Safety

The core objective of this collaboration is to make tangible progress in the scientific methods used to evaluate artificial intelligence systems. As highlighted, the partnership will move beyond merely developing policy frameworks to actively creating practical instruments for assessment. This focus on evaluation sciences is deemed increasingly important given the rapid advancements in AI.
Joint Development of Benchmarks and Tools

Under the memorandum of understanding, IMDA and Microsoft will work together to build specific benchmarks and tools. These will serve as standardized measures and practical instruments for assessing the safety and security aspects of AI. While the specific nature of these benchmarks and tools is not detailed, their creation is central to the partnership's aim of enhancing AI governance.
Addressing Emerging Risks and Building Trust

Natasha Crampton, Microsoft's Chief Responsible AI Officer, noted that the partnership intends to leverage government insights and Microsoft's technical expertise to strengthen AI evaluation. This approach is designed to address emerging risks associated with advanced AI systems and to cultivate greater trust in their deployment. The development of these evaluation methods is a key component in achieving these goals.

Reading history

Hide →

This history is cleared when the page is closed.

What benchmarks and tools will be developed under this AI governance partnership?

Loading

Was this article helpful?

Helpful

Not helpful

Report an issue

2

0

0

Share with friends