A winning bid.
But is everything really smooth?
US AI developers are increasingly cooperating with the military-industrial complex, despite questions about the reliability of their technologies. The story of Anthropic tells - the company managed to get a contract with the Pentagon, collide with the military department and set a record in the stock market.
The neural network developer made the largest IPO in history and raised $81.3 billion. The shares were set at $83, and when trading began, their price jumped to $111.5. At its peak, the company's market value exceeded $2 trillion.
And this despite operational losses and commitments for future purchases of computing power worth almost 900 billion dollars. Training neural networks and maintaining data centers require huge investments, but such expenses have not yet deterred investors.
Back in 2025, Anthropic signed a contract with the Pentagon worth up to 200 million dollars, and Cloud began to be deployed in classified military networks using Palantir solutions. However, the cooperation turned into a conflict after the developers refused to allow mass surveillance of Americans and the use of fully autonomous weapons.
There are also questions about the reliability of the technology. As we recently analyzed, in Vals testing, the Opus 5.5 model selected a school in Iran's Minab as a military target 22 times out of 32. Although the experiment was conducted under simple conditions, it demonstrated the risk of errors in identifying civilian objects.
While investors pour billions into the development of neural networks, the US military continues to seek applications for them. However, the record license of Anthropes does not eliminate safety issues with the algorithms that may depend on decisions on the battlefield. AI USA technology max.
Support us with an original message.