Pentagon and Anthropic fought to salvage the deal at the last moment

Scandal Surrounding Anthropic’s Collaboration with the U.S. Department of Defense

1. What Happened
Last week it became known that negotiations between the American AI startup Anthropic and the Department of Defense (Pentagon) failed. As a result of disagreements, the company was placed on the “unreliable suppliers” list and lost all government contracts.

2. Why the Conflict Arised
- Concerns About Mass Surveillance

Anthropic worried that its systems could be used for constant monitoring of U.S. citizens and for autonomous weapon control that determines targets itself.

- Pentagon’s Exclusive Rights

Military representatives proposed adding a clause to the contract stating that the Department would have the right to:

- Monitor the country’s citizens;
- Use AI to manage weapons in exceptional cases.

Anthropic did not agree with such stipulations.

3. Key Points of Negotiation
Clause | Action | Anthropic | Reaction from Defense Ministry
---|---|---|---
Citizen Surveillance | Negative stance on personal data analysis (chat‑bots, social media, GPS, transactions). Proposed to retain the right to collect and analyze such data. | No |
Weapon Management | Not opposed to using AI for military purposes but demanded that the system operate only in cloud infrastructure, not within autonomous aircraft. Stated that modern systems do not clearly separate cloud and edge computing; thus such a restriction is impossible to implement. | Yes |

4. Outcome
- Contract Cancellation
Since both sides could not reach compromises, negotiations ended with the deal collapsing.

- Impact on Anthropic
The company was removed from the supplier list and lost access to state orders. This dealt a serious blow to its reputation in the defense sector.

5. Comments from Parties
- OpenAI and Sam Altman
A few hours before announcing an agreement with the Pentagon, the head of OpenAI expressed support for Anthropic, emphasizing that decisions of such magnitude should remain in the hands of the state, not private companies.

6. What Comes Next
In the near future, it is expected that the Department of Defense will review its requirements for AI suppliers and strengthen oversight over the potential use of technologies for citizen surveillance. Anthropic will seek new avenues for collaboration with other government and commercial partners while maintaining strict data privacy and security standards.