The Pentagon warns of risks at Anthropic by Chinese employees

The U.S. Department of Defense warned of national security risks linked to the artificial intelligence company Anthropic.

The concern is due to the presence of employees of Chinese origin in key roles in the development of advanced models.

The warning was formalized by the Pentagon's Undersecretary, Emil Michael, in a court filing.

Suspicions of possible influence of the Chinese regime

The Pentagon warned that workers linked to the People's Republic of China could be subject to that country's intelligence legislation

.

Under these regulations, Chinese citizens may be required to collaborate with state intelligence services.

This, according to Washington, increases the risk of sensitive information being leaked.

Anthropic was identified as a supply chain risk In

early March, the Pentagon classified Anthropic as a potential

supply chain risk.

In response, the company took legal action to reverse that decision.

The company seeks that Justice: annul

• the classification

,

• suspend its application

,

• force the Government to continue using its services

Differences with other companies in the sector

.

The Department of Defense maintained that other technology companies also employ foreign talent, but they have

greater security guarantees.

In the case of Anthropic, the authorities consider that the level of risk is different.

This marks a point of tension between the Government and one of the leading firms in artificial intelligence.

The weight of foreign talent in artificial intelligence The

debate reflects a structural reality of the sector.

Cited data indicate that between 38% and 40% of the elite talent in artificial intelligence in the United States is of Chinese origin.

Dependence on international talent is key to maintaining technological leadership, but it also creates security challenges.

The

United States continues to use its services

Despite warnings, the Pentagon continues

to use Anthropic products.

It even considers extending the deadlines to stop working with the company, depending on the development of the judicial dispute.

This highlights the difficulty of rapidly replacing critical capabilities in artificial intelligence.

The defense of Anthropic

The company maintains that it has advanced systems to mitigate internal risks

.

Its measures include:

• compartmentalization of projects

,

• internal audits

,

• access control

.

In addition, in 2023, it became one of the first companies in the sector to formally collaborate with the Pentagon

.

He also claimed to have detected and blocked cyberespionage attempts linked to China.

A key debate for the future of AI The

case focuses on a strategic discussion: how to balance access to global talent with the protection of national security.

Industry experts warn that the risk of infiltration is real, but it is also difficult to manage without affecting competitiveness.

A decision with geopolitical impact

The outcome of the planned court hearing will define the future of the relationship between the US Government

and Anthropic.

It will also set a precedent for how artificial intelligence is regulated in a context of global competition with China.

The tension between innovation and security is consolidated as one of the main challenges of the 21st century.