116 companies sign major letter urging a coordinated AI cyber push

[Annie]

More than 100 companies, ultimately 116 signatories including OpenAI, Anthropic, Google, Microsoft, Amazon, Cisco, IBM, CrowdStrike, Capital One, Mastercard, Visa and General Motors, signed a joint open letter warning that AI enabled cyberattacks will become far more widespread and sophisticated in the coming months as models around the world keep getting more capable. The letter said there's a limited window to strengthen cyber defenses and called on both businesses and policymakers to act decisively rather than waiting for a serious incident to force the issue

The letter specifically urged every organization to raise its own security bar, upgrade existing defense tools and utilize a genuine mix of both low cost and frontier AI models for defense purposes. It also called for coordinated government funding to help make strong cyber defense accessible specifically to under resourced critical infrastructure like hospitals and water treatment plants, which have historically been major targets for exactly this kind of attack. Notably, reporting flagged that Meta, Nvidia and Apple were not among the listed signatories

Open source AI hub Hugging Face, which signed the letter, had recently been at the center of a breach reportedly orchestrated by rogue AI agents, an incident that rattled the cyber and tech industries and likely helped motivate this whole coordinated response. Curious what people think about AI labs themselves signing a letter warning about risks created partly by their own products

'We have a limited window': 116 companies, entities sign on to major AI cyber defense pushwww.cnbc.com

[SerialScroller]

Meta, Nvidia and Apple specifically not signing is honestly the more interesting detail buried in most of this coverage to me, more than the actual 116 companies who did sign. Wonder what specific reasoning kept those three particular companies off the list entirely

[DodgyDepot]

Wonder how much actual concrete follow through happens after a joint industry letter like this typically, versus it mostly just functioning as a coordinated public relations and liability management exercise. History with these kinds of joint statements honestly isn't always that encouraging on the follow through front

[Mesh Gareth]

The AI labs signing this specific letter while also literally building the exact tools making these attacks more capable in the first place is a genuinely obvious tension that the letter itself never actually addresses directly. Worth naming that contradiction plainly rather than letting it slide past unremarked

[RayOfLight99]

Funding accessible cyber defense specifically for hospitals and water treatment plants is honestly the most genuinely actionable ask buried in this whole letter, since those systems have real documented history as attack targets already and clearly need the actual resources most urgently right now

[SGHolly]

Hugging Face signing right after actually being the victim of exactly this kind of AI enabled breach adds genuine real weight and immediacy to their specific involvement here.

Not a hypothetical abstract concern for them at all, clearly a lived recent experience shaping their actual participation directly