UK Media: Anthropic Sues Pentagon but Deploys Offensive AI for NSA
Key facts
- UK Media: Anthropic Sues Pentagon but Deploys Offensive AI for NSA
- The Financial Times reports that AI startup Anthropic, while in litigation with the Pentagon, has deployed engineers inside the US National Security Agency (NSA) to assist with an offensive cyber AI model called 'Claude Mythos'.
- Source: PR Times
- Date: June 5, 2026
Direct answer
The Financial Times reports that AI startup Anthropic, while in litigation with the Pentagon, has deployed engineers inside the US National Security Agency (NSA) to assist with an offensive cyber AI model called 'Claude Mythos'.
- Citation
- UK Media: Anthropic Sues Pentagon but Deploys Offensive AI for NSA (June 5, 2026), PR Times
- Source
- PR Times
- Date
- June 5, 2026
The Financial Times reports that AI startup Anthropic, while in litigation with the Pentagon, has deployed engineers inside the US National Security Agency (NSA) to assist with an offensive cyber AI model called 'Claude Mythos'.
📋 Article Processing Timeline
- 📰 Published: June 5, 2026 at 12:18
- 🔍 Collected: June 5, 2026 at 12:27 (9 min after Published)
- 🤖 AI Analyzed: June 6, 2026 at 15:16 (26h 48m after Collected)
(Central News Agency, London, 4th Combined Foreign Reports) The British Financial Times has revealed that despite being in litigation with the Pentagon, US artificial intelligence startup Anthropic has deployed engineers inside the US National Security Agency (NSA) to deploy an AI model called 'Claude Mythos' for offensive cyber operations.
The report, citing two informed sources, stated that San Francisco-based Anthropic has stationed six employees inside the NSA, working as so-called 'forward-deployed engineers' to help guide the use of this powerful AI and customize the model for specific purposes.
It is unclear whether Anthropic's engineers are also assisting the NSA in executing actual cyberattack missions. However, one informed source indicated that Mythos would be very useful for infiltrating the network systems of countries like China or Iran.
A person close to Anthropic said, 'Attack is the best defense,' and believes that America's adversaries are likely building their own AI attack technologies.
Currently, Anthropic is in a lawsuit with the Pentagon, and the NSA itself is part of the Pentagon system. The crux of the lawsuit includes the ways in which Anthropic's technology is used for combat.
Anthropic is attempting to restrict the US government from using its AI models for mass surveillance of American citizens and for autonomous lethal drones. This led the Pentagon to designate the company as a 'supply chain risk,' the first time a US company has received such a designation. Anthropic has sued over this.
If Anthropic loses the lawsuit, it may be forced to terminate contracts with some agencies that provide services to the US military.
The launch of Mythos by Anthropic has raised concerns among governments, financial institutions, and IT companies about the AI model, particularly its ability to detect and exploit software vulnerabilities.
Earlier this week, Anthropic announced it would provide Mythos to 150 institutions in 15 countries, significantly expanding its availability beyond just the US and UK. When Mythos was launched in April this year, it was only open to a select few specific institutions in the US. (Translation: Chen Yiwei) 1150605
The report, citing two informed sources, stated that San Francisco-based Anthropic has stationed six employees inside the NSA, working as so-called 'forward-deployed engineers' to help guide the use of this powerful AI and customize the model for specific purposes.
It is unclear whether Anthropic's engineers are also assisting the NSA in executing actual cyberattack missions. However, one informed source indicated that Mythos would be very useful for infiltrating the network systems of countries like China or Iran.
A person close to Anthropic said, 'Attack is the best defense,' and believes that America's adversaries are likely building their own AI attack technologies.
Currently, Anthropic is in a lawsuit with the Pentagon, and the NSA itself is part of the Pentagon system. The crux of the lawsuit includes the ways in which Anthropic's technology is used for combat.
Anthropic is attempting to restrict the US government from using its AI models for mass surveillance of American citizens and for autonomous lethal drones. This led the Pentagon to designate the company as a 'supply chain risk,' the first time a US company has received such a designation. Anthropic has sued over this.
If Anthropic loses the lawsuit, it may be forced to terminate contracts with some agencies that provide services to the US military.
The launch of Mythos by Anthropic has raised concerns among governments, financial institutions, and IT companies about the AI model, particularly its ability to detect and exploit software vulnerabilities.
Earlier this week, Anthropic announced it would provide Mythos to 150 institutions in 15 countries, significantly expanding its availability beyond just the US and UK. When Mythos was launched in April this year, it was only open to a select few specific institutions in the US. (Translation: Chen Yiwei) 1150605
FAQ
What are the key facts in this article?
The Financial Times reports that AI startup Anthropic, while in litigation with the Pentagon, has deployed engineers inside the US National Security Agency (NSA) to assist with an offensive cyber AI model called 'Claude Mythos'.
What is the direct answer?
The Financial Times reports that AI startup Anthropic, while in litigation with the Pentagon, has deployed engineers inside the US National Security Agency (NSA) to assist with an offensive cyber AI model called 'Claude Mythos'.
What is the source and date?
PR Times: https://www.cna.com.tw/news/aopl/202606050065.aspx | June 5, 2026