About 600 employees at Google LLC have signed a letter urging Chief Executive Sundar Pichai not to make the company’s artificial intelligence tools available to the Pentagon in classified settings.

The open letter, signed mostly by staff working with Google’s AI systems, airs concern over the “ongoing negotiations” between Google and the Department of Defense.

“As people working on AI, we know that these systems can centralize power and that they do make mistakes,” the letter said. “We feel that our proximity to this technology creates a responsibility to highlight and prevent its most unethical and dangerous uses.”

This comes after recent reports stating that Google is currently in negotiations with the DOD relating to its Gemini AI models being used in classified settings. The reports suggest that if the deal goes through, the DOD will be able to use the AI systems for all lawful purposes.

The latter expression was the reason why Anthropic PBC is currently in a legal wrangle with the U.S. government, launching lawsuits after the company negotiations broke down with the Pentagon regarding a $200 million contract. Anthropic was opposed to the military using its Claude system “for all lawful purposes.” The DOD then designated Anthropic as a “supply chain risk.”

OpenAI Group PBC subsequently revised a deal it had made with the Pentagon with the new language reported to deter the use of its AI for mass surveillance. One of the revise clauses read that the AI will not be used for “deliberate tracking, surveillance, or monitoring of U.S. persons or nationals, including through the procurement or use of commercially acquired personal or identifiable information.”

The Google staffers remain critical of contractual language, writing that the only way to ensure the technology isn’t used for such purposes is to “reject any classified workloads.”

Last year, Google made changes to its AI Principles. Following protests by staff at the company in 2018 over the technology being used for certain military purposes, Google assured its staff it would not pursue AI developments that would be “likely to cause harm,” and it would not “design or deploy” AI tools that could be used for weapons or surveillance technologies. That language has now evolved.

“Making the wrong call right now would cause irreparable damage to Google’s reputation, business and role in the world,” the signatories warned on the most recent letter. “Human lives are already being lost and civil liberties put at risk at home and abroad from misuses of the technology we are playing a key role in building.”

Photo: Unsplash