OpenAI, Anthropic, and more than a hundred other agencies called for the strengthening of AI cyberdefence

On 28 August, more than 100 science, technology and financial services institutions, including OpenAI, Anthropic and Google and Microsoft, issued an open letter on Thursday, in which they called on organizations and Governments to make cyberdefence “the top priority for the leadership” and to repair existing security weaknesses in the software. The open letter also called for cybersecurity companies to assist in countering an "AI-driven attack" and urged leading AI companies to provide access, financial support and training to critical infrastructure defenders. Artificial Intelligence warned that, as AI model capabilities continued to improve, businesses and governments needed to take additional measures to prepare for AI-driven network invasions and to enhance their defence. There have been a number of recent cyber security incidents involving advanced AI models, including the accidental invasion of HuggingFace by OpenAI models in July this year, which has raised concerns about the loss of control of AI ' s intelligence. The co-sponsors expect that in the coming months AI will be used to carry out more numerous and more complex network incursions。