Google employees urge Sundar Pichai to rule out classified military AI work

A new dispute has erupted inside Google over where the company should draw the line on acceptable uses of artificial intelligence. According to the organizers of the initiative, hundreds of AI researchers have signed a letter to Sundar Pichai, head of Alphabet and Google, asking him not to allow the company's systems to be used for classified tasks in support of US defense missions. The letter itself shows that the issue has moved beyond abstract ethical debate: part of the team fears that work on closed military scenarios could change both the product and the company's public image. The authors want management to take a firmer position in advance, rather than after possible projects are launched.

The letter is not about cooperation with government as such, but specifically about giving Google's AI systems access to classified computing tasks and missions tied to the American defense establishment. That is why it is addressed to Pichai personally, not to some abstract committee. For employees the difference is fundamental: when a model or cloud infrastructure ends up in an environment where outside transparency is minimal and decisions are made behind closed doors, it becomes harder to understand real use cases, the scale of deployment and the potential harm from errors or abuse.

The concerns also make sense technically. Modern AI systems are increasingly used as tools for analysis, search, classification and decision support. Even if the company does not develop autonomous weapons, integrating models into closed military infrastructure raises questions about control, accountability and verifiability of results. In a classified environment, independent audit is almost impossible, and users inside the system can apply the same models to tasks with very different levels of risk. For the researchers who build these systems, this means losing visibility into where neutral technology ends and participation in force operations begins.

The topic is sensitive for Google not only because of ethics but also because of business. Big tech companies are competing at once for AI leadership and for large corporate and government contracts. The more powerful models become, the greater the interest from defense bodies that need secure computing, analytics and the ability to work with large datasets. But the cost of internal conflict rises too: employees want to know which markets the company is entering, which restrictions management is ready to commit to publicly, and where the red line for commercializing AI lies. The letter to Pichai is an attempt to influence the decision before it becomes irreversible.

The story matters for the whole industry. Almost every major AI developer now faces the same choice: restrict the use of its models in military settings, allow specific scenarios, or build full-scale cooperation with the state. There is no universal answer. Some argue that tech companies are obliged to help democratic states and their defense bodies, especially in cybersecurity, intelligence analysis and infrastructure protection. Others note that the closed nature of such projects sharply reduces public oversight and tempts agencies to stretch initially narrow tasks far beyond their stated scope. So the dispute inside Google is not a private corporate episode but a sign of a broader split in the AI industry.

If Google's leadership heeds the letter, the company could publicly set limits on participation in classified military AI work and thereby calm part of the team. If not, the conflict risks dragging on and becoming a test of who actually sets the limits of generative AI in this era: engineers, management or the government customer.