The UK and US have signed a groundbreaking agreement to collaborate on testing advanced artificial intelligence (AI). Under the terms of the accord, the two countries will jointly develop rigorous methods to assess the safety of AI tools and the systems that support them. It represents the first bilateral agreement of its kind and is seen as a major step forward in addressing concerns around the development and use of AI. UK tech minister Michelle Donelan labelled it “the defining technology challenge of our generation”.
Donelan argued that ensuring the secure advancement of AI is a shared global issue that can only be addressed through cooperation among nations. The agreement builds on commitments made at the AI Safety Summit at Bletchley Park in November 2023, which saw the UK and US establish AI Safety Institutes to evaluate AI systems that are both open and closed-source. Attendees included tech industry luminaries Demis Hassabis and Elon Musk, as well as the CEO of OpenAI, Sam Altman.
While the safety of AI has come into focus following that summit, competition and technological development continue. The fact that many of the leading players in the AI arena, including ChatGPT, Gemini, and Claude, originate from the United States has meant that regulators have yet to impede or challenge what these companies are doing. The European Union’s upcoming AI Act, however, will require regulatory approval and will mandate more transparency around AI risks and data use. Recently, OpenAI opted not to release a voice cloning tool due to the serious risks it held, especially in an election year.
Concerns over AI’s existential risk are sometimes overblown, according to University of Oxford professor Sir Nigel Shadbolt. While AI is capable of being weaponised like nuclear, chemical, and biological sciences, he believes that “sometimes fears around AI’s existential risk are a bit overblown”. The U.S. Commerce Secretary Gina Raimondo believes this agreement will provide governments with greater visibility into the operation of AI systems, helping them provide more effective guidance
Read the full article from The BBC here: Read More