OpenAI, Anthropic and Google — joined by more than 100 other companies — have co-signed an open letter urging industry players and governments to undertake a “global cybersecurity offensive” against AI-driven cyberattacks. The signatories warn that such attacks are likely to become more intense and effective in the coming months.
Why they are sounding the alarm
The core concern is that advanced artificial intelligence models can significantly enhance the capabilities of actors carrying out cyberattacks. At the same time, the letter notes a pragmatic response: the same AI tools can and should be used to strengthen defenses before these capabilities are weaponized by malicious actors.
Recommended measures
The open letter outlines several specific steps, including:
- Treating cybersecurity as a priority for companies and public service providers.
- Patching all existing high-risk security vulnerabilities as a matter of urgency.
- Updating and modernizing legacy systems.
- Using the most powerful artificial intelligence models to reinforce defensive measures and address the most complex security problems.
- Encouraging cybersecurity firms to continuously test their defenses against the capabilities of state-of-the-art AI models.
Who backed the initiative?
The letter is signed by multiple well-known technology organizations, including OpenAI, Anthropic and Google, together with more than 100 additional companies. Technology media outlets such as Gizmodo have reported on the recommendations and warnings contained in the letter.
Why this matters
The appeal highlights a paradoxical situation: artificial intelligence can both enable more sophisticated attacks and provide stronger means of defense. The signatories argue that, without coordinated and swift action, AI-driven attacks could threaten critical infrastructure and essential public services. The proposed measures aim to ensure defensive capabilities keep pace with rapidly advancing offensive techniques.
Next steps
The authors of the letter call on governments and industry to respond quickly and in a coordinated manner by closing high-risk vulnerabilities, updating systems, and deploying AI-based defensive tools. They also emphasize the need for continuous testing to ensure defenses remain resilient against the latest and most advanced attack methods.
