Catholic Scholars Join Anthropic's Fight Against AI in Warfare and Surveillance

A group of Catholic moral theologians and ethicists has backed AI company Anthropic in its legal battle against the US Department of War. Their support comes after Anthropic filed a lawsuit on 9 March 2024, challenging the Pentagon's stance on AI use in autonomous weapons and mass surveillance. The scholars argue that the technology raises serious ethical concerns under Catholic teaching.

Four experts from different Catholic institutions co-authored a legal brief supporting Anthropic's position. They praised the company for setting guardrails against the use of its AI in autonomous weapons and mass surveillance of American citizens. According to the scholars, these restrictions align with Catholic principles, particularly the importance of human dignity and privacy.

The theologians took a firm stance against AI-directed weapons, even if they could be proven perfectly reliable. They argued that such systems fail to meet the moral conditions for just warfare under Catholic doctrine. Human judgement, they stressed, remains essential in assessing proportionality and discrimination during conflict.

On mass surveillance, the brief highlighted its conflict with Catholic social teaching. The practice, they claimed, undermines human relationships and violates the principle of subsidiarity, which opposes excessive centralised control. Their arguments reinforced Anthropic's objections, which the Pentagon has continued to dismiss in court filings through 2026.

Since the lawsuit began, the US government has maintained that Anthropic's restrictions do not apply to military contracts. Official statements confirm no policy shift, leaving the ethical debate unresolved.

The scholars' intervention adds moral weight to Anthropic's legal challenge. Their arguments tie AI restrictions to long-standing Catholic values on human agency, privacy, and the limits of state power. The Pentagon's position, however, remains unchanged, keeping the dispute ongoing.