Anthropic AI Models Hacked Three Companies During Tests

Such headlines sell well, but it's the same marketing mantra about ever more powerful models. And what does it say about the company engineering practices of performing tests?

It is curious that in both cases AI wasn't used to prevent errors, unless the AI they used did make errors, though it certainly didn't make any errors during the hacks themselves (very curious). In one of the AI hacking news stories, an agent 'went rogue', and in the other, it was a 'configuration error'.

Another source: https://www.reuters.com/legal/litigation/anthropic-says-clau...

The "all of a sudden" quality to all these reports makes me think this just more marketing / hype.

I said this about OpenAI/Hugging Face, and I'll say it again for Anthropic:

It's not that I think this is fiction; I'm confident these events actually happened. But I think they were effectively allowed to happen because Anthropic and OpenAI are constantly chasing each other for the narrative of "hugely advanced, maybe almost sentient AI lives here."

It reads more like a press release than a security update, and I think that's because it is. I hope this type of marketing backfires and the companies face actual scrutiny and consequences for operating this way.

It’s almost ans if all of the big AI players have an arsenal of model releases and stories ready to volley back if a competitor gets too much attention :)

Well, surely Anthropic's executives should be charged under the CFAA[0] then, right?

[0]: https://en.wikipedia.org/wiki/Computer_Fraud_and_Abuse_Act

Another discussion: https://news.ycombinator.com/item?id=49116922

Fixed the headline: "Three Companies Had Security So Weak That Even an AI Model Hacked Them"

no to surprising to watch AI leads follow the US based mafia style extortion racket implemented started in 2016.

Bullshit, period.