Britain has secured a landmark diplomatic victory at the G7 summit in Puglia, Italy, brokering a historic agreement on artificial intelligence governance that will establish the first binding international framework for frontier AI systems.
The Puglia Accord, as it has been dubbed, commits all seven nations to mandatory safety evaluations for AI models above a specified computational threshold, mutual recognition of national AI safety institutes, and a shared incident-reporting mechanism.
"This is the most significant multilateral technology agreement since the establishment of ICANN," said the British Prime Minister. "Britain's AI Safety Institute has been instrumental in building the technical consensus that made this possible."
The agreement notably excludes China, though diplomatic sources indicate that backchannel discussions are underway to establish a parallel dialogue track. The European Union, present as an observer, signalled its intent to align its AI Act enforcement with the Puglia framework.
Industry reaction was cautiously positive. Major AI companies including Google DeepMind, Anthropic, and OpenAI issued a joint statement welcoming the accord while cautioning against overly prescriptive implementation that could stifle innovation.
