Google DeepMind engineer resigns over military AI contracts

The catalyst for resignation

Sofia Liguori, a research engineer at Google DeepMind, officially resigned from the company in an open letter dated October 20, 2026. She spent more than three years at the company, where she worked on research involving optimal power flow, energy, and sustainability. Her departure follows a period of leave that began after Google signed a classified agreement with the US Department of War in April.

Liguori identified this specific classified deal as the primary reason for her exit. She linked the new agreement to Project Nimbus, which is a cloud computing arrangement between Google and the Israeli government. She argues that these types of contracts allow AI to be integrated into military and surveillance systems without any meaningful restrictions on their final use.

In her public statement, Liguori expressed a loss of faith in the company's direction. She stated, "I no longer believe that our technology will have a net positive impact on society." Her resignation serves as a public critique of how large tech firms balance commercial interests with the potential consequences of their software in warfare.

Concerns over oversight

A central theme of Liguori's critique is the lack of visibility into how customers actually deploy AI technology. She expressed concern that companies cannot uphold ethical commitments when they have limited control over the end-use of their products. She specifically pointed to Google's relationships with Lockheed Martin and several US agencies, including the Department of Homeland Security and Customs and Border Protection.

Liguori argued that the current business model for these contracts leaves the company with no monitoring power. She characterized the situation as a standard being set where AI is integrated into military systems while imposing no rules on permissible uses. This, she suggests, creates a situation where the technology can be used in ways that may contradict the original intentions of the developers.

The engineer also challenged the idea that technology is a neutral tool. She argued that developers have a responsibility to assess a tool based on the context in which it will be used rather than just its technical capabilities. She noted that while her own work was task-specific, the general-purpose and agentic AI models being developed have much broader implications for society.

Allegations of intimidation

Beyond the specific contracts, Liguori alleged that Google DeepMind maintains a workplace culture that discourages employees from speaking out. She described a corporate environment where military contracts are treated as off-limits topics that staff should avoid raising. She claimed that even teams whose work might have implications for these deals are pressured to remain silent.

According to Liguori, she faced meetings with Human Resources over alleged internal policy breaches when she attempted to challenge these boundaries. She characterized these interactions as a form of intimidation rather than a substantive disciplinary process. She called for a rejection of what she termed a "corporate culture of censorship and intimidation" regarding the company's defense partnerships.

In response to these types of allegations, a Google spokesperson told the BBC that the company follows a consistent process for reviewing public sector contracts. They stated that they adhere to acceptable use policies and AI principles. The spokesperson also noted that the company has dedicated channels for employees to raise concerns about ethics or misconduct without fear of retaliation.

The path forward for workers

Despite her resignation, Liguori did not want her departure to be an isolated act of protest. She argued that collective employee action is more effective than individual resignations for influencing how large AI companies operate. She encouraged remaining employees to organize to exert influence over the direction of AI development and deployment.

She directed current workers toward several specific organizing efforts, including groups like Googlers Against ICE and the Coalition of Concerned AI Staff. Her goals for these organizations include demanding enforceable restrictions on surveillance applications and greater transparency regarding contracts. She also called for the right of employees to abstain from specific work on ethical grounds and for stronger sustainability commitments from the company.

Liguori's letter serves as a call to action for the broader tech workforce to consider the concentration of political and technological influence. She believes that researchers and engineers still retain considerable influence over the development of AI. Her message emphasizes that responsibility does not end with the technical completion of a system but extends to who controls it and who benefits from its use.