AI and Cyber: Over 100 Companies Warn About Infrastructure

More than 100 companies, initiated by OpenAI, are calling for a wake-up call in response to cyber threats amplified by AI. Their alert comes with specific demands directed at boards of directors, governments, and AI providers, as U.S. agencies report observing offensive uses of AI against industrial systems.

Offensive Uses of AI Already Observed According to U.S. Agencies

A joint warning published in mid-August by the NSA, CISA, and FBI indicates that attackers are already exploiting AI to write scripts targeting industrial control systems, notably the Siemens S7. These attacks affect sectors such as energy, water, chemicals, and manufacturing in the United States.

Measures Requested from Leaders, States, and AI Providers

The signatories of the open letter urge companies to elevate cybersecurity to the level of boards of directors. They call on governments to increase funding and coordination in cybersecurity, and invite AI providers to offer affordable security tools for underfunded organizations. They encourage the deployment of AI tools while defenders still have an advantage, believing that current advancements in AI can already address weaknesses that have accumulated over the years. Long-standing vulnerabilities, such as unpatched software and weak authentication, are deemed a priority to address.

A Coalition of Over 100 Stakeholders and Priority Targets

OpenAI's open letter brings together more than a hundred companies, including Microsoft, Google, AWS, Anthropic, Cisco, CrowdStrike, Deutsche Telekom, SAP, and Mastercard. Cyberattacks using AI are expected to become significantly more frequent and complex. Hospitals, water networks, and other critical infrastructures are among the sectors considered most vulnerable.