Google yields on “any lawful use” of artificial intelligence for the Pentagon, while employees protest.
The Mountain View tech giant reportedly signed a classified agreement allowing the United States Department of Defense to use its AI models for “any lawful government purpose,” according to The Information. The agreement was revealed less than a day after Google employees asked CEO Sundar Pichai to prevent the Pentagon from using its AI, fearing it could be employed in “inhumane or extremely harmful ways.”
As The Verge points out, if the agreement is confirmed, Google would join OpenAI and xAI, which have signed classified AI agreements with the US government. Anthropic was also among these, before being blacklisted by the Pentagon for refusing to comply with the Department of Defense’s requests to remove restrictions related to weapons and surveillance from its AI models.
All the details.
THE AGREEMENT BETWEEN GOOGLE AND THE PENTAGON ON AI
The agreement would allow the Pentagon to use Google’s AI systems without the Alphabet-owned company exercising direct control over operational methods. The contract would establish that Google has “no right of control or veto over the government’s legitimate operational decisions.”
Although it includes indications that AI should not be used for domestic mass surveillance or autonomous weapons “without adequate human oversight and control,” such limits would not be legally binding. The agreement also provides for the possibility that Google collaborates on adjusting safety settings and filters at the government’s request.
THE POSITION OF THE GIANT LED BY SUNDAR PICHAI
“We are proud to be part of a broad consortium of leading AI labs and technology and cloud companies providing AI services and infrastructure in support of national security,” a Google spokesperson said in a statement to The Information, clarifying that this is an amendment to a pre-existing government agreement.
THE OPEN LETTER FROM BIG G EMPLOYEES
The agreement emerged shortly after more than 560 Google employees signed an open letter to CEO Sundar Pichai urging him to refuse to allow the US government to use its artificial intelligence technology for classified military operations.
“We want AI to benefit humanity, rather than be used in inhumane or extremely harmful ways. This includes lethal autonomous weapons and mass surveillance, but extends beyond,” they wrote.
“The only way to ensure that Google is not associated with this is to refuse any classified workload,” the letter emphasizes.
AFTER THE AGREEMENTS SIGNED BY OPENAI AND XAI
If confirmed, the agreement would place Google among the leading tech companies collaborating with the US government on classified AI projects, alongside OpenAI and xAI. Anthropic had also signed a similar agreement before being excluded by the Pentagon after refusing to remove certain restrictions related to military use and surveillance.
THE ANTHROPIC CASE
The AI company led by Dario Amodei refused to authorize the use of the AI-based chatbot Claude for US surveillance or autonomous weapons, due to safety and ethical concerns. In response, the Pentagon applied a supply chain risk label to Anthropic.
PREVIOUS PROTESTS BY GOOGLE EMPLOYEES
Finally, the issue of military use of technologies is not new for the Mountain View giant.
In the summer of 2018, Google abandoned the Pentagon’s Maven project. The program focused on using computer vision software to automatically analyze footage collected by US military drones. Google was forced into the decision by employee protests demanding the company stay out of the war business.
Also in 2018, company employees signed an open letter urging Google to cancel the Dragonfly project regarding a censored search engine for the Chinese market that would have allowed state surveillance.
In recent years, however, Google has progressively reopened collaboration with the military sector. Last year it removed some restrictions on the use of the technology and in December signed an agreement with the Department of Defense for the deployment of the Gemini AI system.
At the same time, the Pentagon reiterated the need to be able to use commercial AI technologies for “all lawful uses,” a formulation that, according to officials, ensures operational flexibility while respecting US military laws and procedures.
