US DoD Ends Anthropic Contract: AI Ethics, Military Use & Global AI Governance Debate

The US Department of Defence (DoD) has abruptly ended its contract with Anthropic, the AI company behind the Claude platform.

• The DoD took the extreme step of labelling Anthropic a ‘Supply Chain Risk’, effectively branding the company as untrustworthy.

The Ethical Boundary

• Firm Ethical Boundaries: Anthropic maintained strict AI safety principles and refused to compromise its ethical standards despite government pressure.
• Rejection of Domestic Surveillance: The company declined requests to deploy its AI systems for large-scale domestic monitoring, citing concerns over privacy, civil liberties and potential misuse.
• Opposition to Autonomous Weapons: Anthropic also refused to integrate its AI into autonomous weapons systems.

The Political Bullying

• Political Labelling and Delegitimisation: After refusing government demands, Anthropic was labelled as “leftist” and “radical” by certain political actors, aiming to discredit the company’s ethical stance
• Allegations of Ignoring National Interest: The company was accused of undermining national security and of not prioritising the state’s strategic interests.
• Narrative of Foreign Alignment: Some critics suggested that resisting government requests indirectly aided rival nations.
• Coercive Power Dynamics: This reflects a broader power struggle in which a state seeks to compel a private technology firm to compromise ethical safeguards in pursuit of perceived geopolitical advantage.

Global Power Play and Hypocrisy

• Safety Commitments vs Military Use: While participating in global AI safety discussions such as the AI Safety Summit at Bletchley Park (UK), the United States is simultaneously exploring military uses of AI.
• Military Use of AI: Reports indicate that AI assistance is already being used in operations against Iran.
  • If a technological frontrunner like the United States deploys lethal autonomous weapons, other powers such as China, India and Japan are likely to follow to avoid strategic disadvantage.
• Authoritarian Misuse Risk: If democratic countries begin using AI for domestic surveillance, authoritarian regimes may justify similar tools to spy on citizens and political opponents. Example: Pegasus spyware scandal.

The Profit vs. Ethics Dilemma

• Ethics vs Market Competition: While Anthropic refused to provide AI services for controversial military uses on ethical grounds, its competitor OpenAI signed a contract with the US Department of Defence to provide similar services.
• Limits of Self-Regulation: This episode highlights that profit-driven private companies may prioritise commercial opportunities over global safety,  indicating the need for stronger public regulation of AI technologies.

Way Forward

• Global AI Treaty: A binding international agreement, similar to the Nuclear Non-Proliferation Treaty (NPT), is needed to regulate the development and use of advanced AI technologies.
• UN-Backed Regulatory Body: An international watchdog under the United Nations is required to monitor states and technology companies, as voluntary self-regulation has proven inadequate.

Check Out UPSC CSE Books

Visit PW Store

Conclusion

Despite the growing need for international AI treaties, the effectiveness of global governance remains uncertain, as powerful states often bypass multilateral institutions and act unilaterally in pursuit of strategic interests.

DOWNLOAD PDF