Google deletes policy against using AI for weapons or surveillance

Google has quietly deleted its pledge not to use AI for weapons or surveillance, a promise that had been in place since 2018.

First spotted by Bloomberg, Google has updated its AI Principles to remove an entire section on artificial intelligence applications it pledged not to pursue. Significantly, Google’s policy had previously stated that it would not design nor deploy AI technology for use in weapons, or in surveillance technology which violates “internationally accepted norms.”

Now it seems that such use cases might not be entirely off the table.

“There’s a global competition taking place for AI leadership within an increasingly complex geopolitical landscape,” read Google’s blog post on Tuesday. “We believe democracies should lead in AI development, guided by core values like freedom, equality, and respect for human rights. And we believe that companies, governments, and organizations sharing these values should work together to create AI that protects people, promotes global growth, and supports national security.”

While Google’s post did concern its AI Principles update, it did not explicitly mention the deletion of its prohibition on AI weapons or surveillance. 

When reached for comment, a Google spokesperson directed Mashable back to the blog post.

“[W]e’re updating the principles for a number of reasons, including the massive changes in AI technology over the years and the ubiquity of the technology, the development of AI principles and frameworks by global governing bodies, and the evolving geopolitical landscape,” said the spokesperson.

Google’s AI Principles listing the “Applications we will not pursue” as of Jan. 30.
Credit: Screenshot: Mashable / Google

Google first published its AI Principles in 2018, following significant employee protests against its work with the U.S. Department of Defense. (The company had already infamously removed “don’t be evil” from its Code of Conduct that same year.) Project Maven aimed to use AI to improve weapon targeting systems, interpreting video information to increase military drones’ accuracy. 

In an open letter that April, thousands of employees expressed a belief that “Google should not be in the business of war,” and requested that the company “draft, publicize and enforce a clear policy stating that neither Google nor its contractors will ever build warfare technology.”

The company’s AI Principles were the result, with Google ultimately not renewing its contract with the Pentagon in 2019. However, it looks as though the tech giant’s attitude toward AI weapons technology may now be changing.

Google’s new attitude toward AI weapons could be an effort to keep up with competitors. Last January, OpenAI amended its own policy to remove a ban on “activity that has high risk of physical harm,” including “weapons development” and “military and warfare.” In a statement to Mashable at the time, an OpenAI spokesperson clarified that this change was to provide clarity concerning “national security use cases.”

“It was not clear whether these beneficial use cases would have been allowed under ‘military’ in our previous policies,” said the spokesperson.

Opening up the possibility of weaponised AI isn’t the only change Google made to its AI Principles. As of Jan. 30, Google’s policy listed seven core objectives for AI applications: “be socially beneficial,” “avoid creating or reinforcing unfair bias,” “be built and tested for safety,” “be accountable to people,” “incorporate privacy design principles,” “uphold high standards of scientific excellence,” and “be made available for uses that accord with these principles.”

Now Google’s revised policy has consolidated this list to just three principles, merely stating that its approach to AI is grounded in “bold innovation,” “responsible development and deployment,” and “collaborative process, together.” The company does specify that this includes adhering to “widely accepted principles of international law and human rights.” Still, any mention of weapons or surveillance is now conspicuously absent.