The UK has debuted what it calls a landmark toolset for artificial intelligence (AI) safety testing.
new offer, announced On Friday (May 10), the country's… AI Safety Instituteis named “Inspection” and is a software library designed for testers such as startups, academics, and And AI developers to international governments assess the specific capabilities of individual AI models and then generate a score based on their results.
Inspect is the first AI safety testing platform overseen by a state-backed institution and released for wider use, according to a news release from the institute.
“As part of the continued drumbeat of UK leadership on AI safety, I have cleared the AI Safety Institute's testing platform – called Inspect. Open source“, said Michelle DonlonUK Secretary of State for Science, Innovation And Technology
“This puts UK intelligence at the heart of global efforts to secure AI, and cements our position as a world leader in this space.”
The announcement comes a little more than a month after the British and US governments pledged to work together on the safe development of AI. Agree to cooperate On the test of the latest AI models.
“AI is evolving rapidly, and both governments recognize the need to act now to ensure a common approach to AI security that can keep pace with the technology's evolving threats.” US Department of Commerce said at the time.
The two countries also agreed to form alliances with other countries to promote AI safety around the world, conduct at least one joint test on a publicly accessible model and “exchange personnel” between the two organizations. with a plan to tap into the collective pool of expertise through exploration. .
Follows through on commitments made in partnership. AI Safety Summit In November last year, world leaders explored the need for global cooperation to tackle the potential risks associated with AI technology.
“This new partnership means that a lot more responsibility will be placed on companies to ensure that their products are safe, reliable and ethical,” the AI ethics evangelist said. Andrew Perry. A global intelligent automation company ABBYY told PYMNTS shortly after the collaboration was announced.
“The tendency of innovators of disruptive technologies is to release products with a 'ship first and fix later' mentality to gain early advantage. For example, while OpenAI ChatGPT's potential Somewhat transparent about the risks, that Released it for widespread commercial use. Despite its harmful effects.”
For all PYMNTS AI coverage, subscribe daily. AI Newsletter.