Behind all this is another question: should tech companies ban things that are legal but they find morally objectionable? The government certainly saw Anthropic’s willingness to play this role unacceptable. On Friday evening, eight hours before the US launched strikes on Tehran, Defense Secretary Pete Hegseth issued harsh comments on X. “Anthropic delivered a master class in arrogance and betrayal,” he wrote, and echoed President Trump’s order for the government to stop working with the AI company after Anthropic demanded its model cloud be stopped from being used for autonomous weapons or mass domestic surveillance. “The War Department should have full, unrestricted access to Anthropic’s models for every lawful purpose,” Hegseth wrote.
But until OpenAI’s full contract reveals more, it’s hard not to see the company sitting on an ideological cliffhanger, promising that it does The Pentagon will proudly use it to do what it sees as the right thing, citing the law as the main basis for what it can do with its technology.
There are three things to look at here. One is whether the positions will be enough for OpenAI’s most important employees. With AI companies spending so heavily on talent, it’s possible that some at OpenAI consider Altman’s justification an unforgivable compromise.
The second is the scorched-earth campaign that Hegseth has promised to wage against Anthropic. Going further than canceling the government’s contract with the company, he announced that it would be classified as a supply chain risk, and “any contractor, supplier, or partner that does business with the United States military may not conduct any business activity with Anthropic.” There is significant debate about whether this is the death blow. legally possibleand anthropic has Said If the threat is followed through it will prosecute. OpenAI has also come to the fore against move.
Finally, how will the Pentagon remove the cloud – the only AI model it actively uses in classified operations, Involved In some Venezuela—while this instead of Attacks against Iran? Hegseth gave the agency six months to do so, during which the Army would phase out OpenAI’s model as well as Elon Musk’s model. xai.
But Claude was reportedly used The attacks on Iran just hours after the sanctions were issued show that it would be easier to phase them out. Even if the months-long feud between Anthropic and the Pentagon is over (which I doubt), we are now looking at the Pentagon’s A.I. acceleration plan With new tensions in the Middle East as the primary testing ground, companies are being pressured to abandon the lines in the sand they once drew.
If you have information to share about how this is unfolding, contact me via Signal (username: jamesodonnell.22).
