Bill The Investor
Bill The Investor|2月 27, 2026 23:22
The most sensational storyline of Anthropic is actually a two legged walk: first, the copyright and "thief shouting catch thief" controversy, and then the direct collision with the Pentagon's hard core. The timeline is roughly as follows: In 2025, Anthropic was sued collectively by writers for training Claude with a massive amount of pirated books. In September, a settlement was reached for a huge sum of $1.5 billion (approximately $3000 per book on average) to destroy illegal data. On the surface, the image of the 'security company' has been damaged. January 2026: Claude is suspected to have been used in the US military's operation to capture Maduro. Anthropic executives inquire with Palantir about his usage, angering the Pentagon. Starting from mid February, the Department of Defense has requested the lifting of Claude's hard restrictions on "mass surveillance of American citizens" and "fully autonomous weapons," threatening to list them as "supply chain risks" and tearing up a $200 million contract. On February 23rd, Anthropic suddenly posted an article accusing DeepSeek, Moonshot, and MiniMax of using 24000 fake accounts to "distill industrial grade" Claude for over 16 million conversations, raising the issue to the level of national security and export control. Musk is the first to step out of the yin-yang: You just lost 1.5 billion yuan in piracy money yourselves, and now you're calling someone else to steal it? February 24-27: The Pentagon issues an ultimatum to tear up contracts without lifting restrictions before 5:01 pm on Friday. Dario Amodei publicly declined: 'I cannot do it in conscience'. Trump directly ordered federal agencies to shut down all Anthropic products within six months and denounced them as' left-wing lunatics'. Comment: Anthropic's actions have indeed turned the banner of "AI security" into a double-edged sword. On the one hand, adhering to principles and rejecting the military's bottomless use has indeed won some moral voice; On the other hand, the high-profile accusation of Chinese companies "stealing technology" in the same week when they were cornered by the Pentagon is too coincidental, and it is difficult not to feel that they are showing loyalty and transferring firepower to Washington. The result was not pleasing to both ends: domestically, it was criticized as a "thief shouting to catch a thief", and the military was directly banned by Trump to take down government orders. Simply put, it may be an impossible triangle for AI companies to simultaneously excel in both "safest" and "most profitable (especially for military orders)". Anthropic has really bumped idealism into the iron plate of reality this time. With a current valuation of 380 billion yuan, the biggest suspense is how it will survive in the midst of business wars and geopolitics. What do you think of the play 'Security Company vs. War Machine'? AI Anthropic Claude
+5
Mentioned
Share To

Timeline

HotFlash

APP

X

Telegram

Facebook

Reddit

CopyLink

Hot Reads