Google has excised from its artificial intelligence principles the commitment to refrain from using the technology as a weapon, according to Bloomberg.
The publication noted that the company’s “AI Principles” previously included a passage promising not to deploy AI applications and technologies “to cause harm, including as weapons.” This wording has now been removed.
In response to a request for comment, Google referred to a blog post.
“We believe democratic countries should lead in AI development, guided by core values like freedom, equality, and respect for human rights. Companies, governments, and organizations sharing these values should collaborate to create AI that protects people, fosters global growth, and supports national security,” emphasized Google’s Senior Vice President James Manyika and DeepMind’s head Demis Hassabis.
Former head of Google’s ethical AI department, Margaret Mitchell, remarked that the removal of the “harm” phrase might indicate the company’s work on technology capable of causing fatalities.
This change is part of a broader policy review among major AI firms. Previously, Meta opened its neural networks to U.S. government agencies and defense contractors, as well as allies. Anthropic took a similar step.
In January, OpenAI entered into an agreement with America’s National Laboratories to apply new AI models for nuclear weapons research.
In August 2024, 200 employees of Google’s AI division, DeepMind, expressed dissatisfaction with the tech giant’s involvement in military initiatives. They called for the termination of contracts with military organizations.
Investments in AI
Meanwhile, Google CEO Sundar Pichai announced plans to invest approximately $75 billion in capital expenditures to expand the artificial intelligence sector.
He believes the funds will accelerate progress in AI innovations and continue to strengthen the company’s core operations.
Capital expenditures are funds spent on acquiring long-term physical or core assets for business operations. It is unknown how much is specifically allocated to artificial intelligence.
Back in January 2025, Meta CEO Mark Zuckerberg promised to maintain high spending on artificial intelligence despite the panic caused by the DeepSeek AI model.