More than 100 technology, AI and cybersecurity companies, led by OpenAI, have signed an open letter warning that AI-enabled cyber attacks could soon become more widespread and sophisticated. The signatories say current security practices are inadequate, particularly for critical infrastructure such as hospitals, water treatment plants and internet services, but the article notes the irony that many of these firms helped create or have long profited from the systems now seen as vulnerable.

The letter calls for wider access to AI-powered defensive tools, continuous testing against frontier AI capabilities, improved threat-intelligence sharing and government funding for essential services. It identifies longstanding weaknesses including unpatched software, poor configuration and weak authentication, while urging AI developers to provide funding, training and responsible access to defensive models. However, it sets no funding amount, deadlines or binding commitments, despite concerns that autonomous AI agents and AI-generated exploit code are already being used to find and exploit vulnerabilities.