- Anthropic CEO Dario Amodei has issued a lengthy internal memo
- It attacks OpenAI’s messaging around its new Pentagon deal
- Anthropic and Claude may still be in talks with the US government
If you thought the debate surrounding AI companies’ dealings with the US military would die down, think again: Anthropic CEO Dario Amodi has quite bluntly accused rival OpenAI of telling “outright lies” about its deals with the Pentagon.
Late last week, Anthropic walked away from a new deal with US intelligence, citing security concerns over the use of artificial intelligence in mass surveillance (especially on domestic citizens) and fully autonomous weapons. In response, US officials and the president himself declared that Anthropic’s AI bot Claude would no longer be used by government agencies.
OpenAI and CEO Sam Altman quickly stepped in, announcing their own deal with the US military that apparently had “more safeguards” than what was offered to Anthropic. Users were far from convinced of the ethics: in the days since, the number of ChatGPT uninstalls has increased significantly, and Claude has quickly climbed the App Store charts.
In an internal memo (via The Information), Amodei’s response has been to question OpenAI’s claims. He calls OpenAI’s overall announcements “false,” and suggests that it contains phrases like “layers of security” that don’t quite hold up. In Amodei’s view, many of the assurances OpenAI has given are “security theater.”
OpenAI “places employees”
Amodei goes on to say that OpenAI is more concerned with “placing employees” than actually securing the use of AI, and questions the caveat in the Pentagon agreement that mentions “all lawful uses” — something that can be rather a gray area when it comes to issues like home surveillance permissions.
The memo also highlights how AI can be tricked and abused – in the most basic way, by simply lying to it about the nature of the data it processes – while underscoring Anthropic’s focus on safety and security. The approaches OpenAI is taking here “mostly don’t work,” says Amodi.
OpenAI chief Sam Altman will almost certainly have more to say, and has already admitted that the first OpenAI announcement was “rushed” and “sloppy”. Meanwhile, the Financial Times reports that Anthropic and Claude may be finding a way back to a deal with the US military after all – although it’s not clear what the terms would be.
The report doesn’t add much about the latest negotiations or how they could potentially affect the OpenAI deal, but it seems that the relationship between Anthropic and the US government might not be completely over just yet.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews and opinions in your feeds. Be sure to click the Follow button!
And of course you can also follow TechRadar on YouTube and TikTok for news, reviews, video unboxings, and get regular updates from us on WhatsApp also.
