Technology

AI firms and cybersecurity leaders warn of potential surge in AI-driven cyberattacks

A joint letter signed by OpenAI, Anthropic, Google, and Microsoft urges organizations to strengthen defenses within months.

Updated Aug 27, 2026 · 21:17 UTC1 reports

The short version

Leading AI developers, cybersecurity firms, and major banks issued a joint open letter warning that AI-enabled cyberattacks could escalate significantly in the coming months.
The letter highlights that current cybersecurity measures are insufficient due to legacy software vulnerabilities, unpatched systems, and weak authentication.
Signatories called on organizations to upgrade legacy infrastructure, urged AI companies to provide responsible model access, and asked governments to fund coordinated defense strategies.

Executives and experts from OpenAI, Anthropic, Google, Microsoft, CrowdStrike, Citi, and Capital One signed an open letter warning of potential AI-driven attacks on hospitals, technology firms, and public infrastructure.[CBS News]
The open letter states that defenders have a limited window, potentially lasting only months, to upgrade cyber defenses using advanced AI tools before attacks become more severe.[CBS News]
A cybersecurity report published by CrowdStrike stated that AI-enabled cyberattacks rose by 89% in 2025 compared to 2024.[CBS News]
The signatories urged frontrunner AI companies to secure their models, fund cybersecurity training, and offer responsible access to defensive technologies.[CBS News]

The exact timeframe and full scale of future AI-enabled cyberattacks remain uncertain, as signatories described the remaining defensive window as an estimate of potentially only months.[CBS News]