A director overseeing Android platform security at Google has stepped away from his role, pointing to the company's expanding work with the US Department of Defense as the key factor. René Mayrhofer detailed his choice in an internal note that later appeared on his personal site.

Mayrhofer described the move as unavoidable once Google agreed to supply its AI models for classified Pentagon tasks. He highlighted how leadership had quietly dropped earlier carbon-neutral targets because of the heavy power needs tied to AI systems. He also criticized new contracts with what he termed the US Ministry of War, noting that broad allowances for any lawful purpose had already clashed with international rules in past cases.

The executive confirmed the letter's accuracy and stressed that strong colleagues remain on his former team. Still, he could no longer support the broader shift toward military collaboration. Google announced the defense agreement in late April, triggering pushback from employees who had asked leaders to avoid such pacts. The company expressed pride in contributing to national security efforts through a group of AI labs while pledging not to support domestic mass surveillance or fully autonomous weapons without human oversight.

'I am a pacifist'

In the note, Mayrhofer referenced Google's 2018 AI guidelines that once barred development of weapons or surveillance tools. Those limits were lifted in an update during February 2025. He joined the firm in 2017 and observed that decision-making had grown less open, with major choices now made at the highest levels without wider discussion.

Mayrhofer is not alone in voicing concerns. Another researcher at Google DeepMind publicly expressed shame over the classified-work agreement. As a self-described pacifist, Mayrhofer said he had long avoided roles supporting offensive military actions. He added that actively contributing to harm was something he refused to accept.

The director, who also teaches at Johannes Kepler University Linz in Austria, warned that the US government had grown less friendly toward academic circles. He cited reports of European universities cautioning staff about US ties and feared the new deal could lead to AI-driven monitoring of EU residents. He concluded that the arrangement effectively positioned Google products against his own community, leaving resignation as the only option.

Mayrhofer noted his relatively secure position as a tenured academic allowed him to prioritize principles over pay. He plans to complete his notice period through August yet will stop any AI-related tasks connected to the defense contract right away. He closed by expressing sadness over the situation and hoping the company might regain its ethical direction.