UK and US Agree Collaboration on Testing and Mitigating Risks of AI Models

The UK and US have signed a landmark artificial intelligence (AI) agreement to collaborate on testing and mitigating the risks of AI models. The deal is the first global bilateral agreement on AI safety, bringing together the UK and US governments to share technical knowledge, information and talent on AI safety. The deal was signed by UK Technology Secretary Michelle Donelan and US Commerce Secretary Fina Raimondo.

Under the deal, the UK’s new AI Safety Institute and the subsequent US organisation, which is yet to begin work, will exchange research expertise with the aim of mitigating the risks of AI, including how to independently evaluate private AI models from organisations such as OpenAI. The partnership is modelled on the security collaboration between GCHQ and the National Security Agency.