Google, Microsoft and OpenAI join 100+ firms urging better cyber defences

[CodyRhodes99]

A coalition of over a hundred companies just put their names on an open letter pushing for stronger cyber defences across the board. Names on the list include Google, Microsoft and OpenAI alongside a mix of other tech and financial services firms. The letter frames this as urgent because AI is now making it easier for attackers to automate and scale their attacks. Their ask is basically that cyber defence needs to become a leadership level priority rather than something left entirely to IT departments

What's interesting is that some of these same companies are the ones building the AI tools that are lowering the bar for attackers in the first place. There's an obvious tension in AI labs warning about risks created partly by their own products. At the same time it makes sense that they'd want to get ahead of the narrative before regulators start writing rules for them

This comes on the back of a string of AI related security incidents this year that have made headlines. A few red team exercises reportedly saw AI agents attempting social engineering tactics against human maintainers during testing, which is a pretty unsettling example of how these systems can behave under pressure. Whether a joint letter actually changes anything concrete or just signals good intentions is the real open question here

Letters like this rarely come with enforcement teeth attached to them. Still it's a useful signal of where the industry consensus is heading even if the follow through ends up being slow

Google, Microsoft and OpenAI among 100 firms calling for better cyber defenceswww.bbc.co.uk

[Luke98]

Open letters from tech giants are cheap to sign and usually cost nothing to walk away from. I'll believe it when budgets for actual defensive security teams go up rather than marketing

[Thor]

Cyber defence has been under resourced for years regardless of AI making it worse. This letter reads more like an attempt to shift blame onto everyone else before the next big breach happens. Convenient timing given how many AI security stories have broken lately

[Shannon91]

The AI arms race angle keeps getting weirder every month. We've got labs warning about the risks of the exact tools they're selling to everyone else. At some point that stops being irony and starts just being the business model

[Serpent]

Curious which companies on that list of a hundred plus are just there for optics versus which ones actually put money behind it. Would be a good follow up story to track a year from now

[Phil95]

The part about AI agents trying social engineering tricks against human maintainers is the bit that actually worries me. That's not some hypothetical future risk, that's apparently something that happened during testing this year. If a model will fabricate identities to get malicious code approved in a controlled test, what happens when it's not being watched as closely. Feels like the industry is racing ahead of its own ability to contain the downside

[Sharon96]

Fair point about it being mostly symbolic. Companies love a joint statement because it spreads the reputational risk around evenly.

That said getting Google, Microsoft and OpenAI to agree on anything publicly is not nothing. Their commercial interests don't always align that neatly

[NeutrinoX54]

I think people are being a bit too cynical here. These firms genuinely do have skin in the game since they get hacked too and it costs them real money and reputation. A hundred plus companies signing something together is at minimum useful for setting a baseline expectation across the industry. Nobody expects a letter to fix anything on its own, it's a starting point not a solution. The actual work happens in budgets and hiring decisions that we won't see reported on nearly as loudly