UK and US to partner on safety testing AI models
The UK has signed an agreement with the US that will see the two countries’ AI Safety Institutes work together to test emerging AI models.
The Memorandum of Understanding will see the two align their scientific approaches and exchange information and personnel, as well as carry out joint testing exercises on AI models.
The announcement follows a commitment made last November, when major AI firms in attendance such as OpenAI and Google DeepMind, agreed to a voluntary scheme that would allow AI safety institutes to evaluate and test new AI models before they were released.
