Google DeepMind staff resign over AI deals with US military, raise safety concerns: Report

According to the BBC report, the departing employees say the military deals go beyond what they expected when they joined the company. They are also worried that existing safety measures and human supervision may not be enough once AI systems are deployed on military networks that are disconnected from the public internet. Despite restrictions in her employment contract, she spoke to the BBC about her reasons for leaving.

"Realising my company is pushing more and more for AI systems in military operations makes me feel ashamed to work at Google," Liguori said In response, a Google spokesperson told the BBC: "We work with public sector organisations worldwide and follow a consistent process for reviewing and entering these contracts, including compliance with our acceptable use policies and AI principles." "We also have dedicated channels for employees to safely raise any concerns - ranging from business ethics to misconduct - and have strict workplace policies in place to prevent retaliation of any kind."

However, as AI technology develops rapidly, concerns are growing over whether these systems are reliable enough for warfare without more testing, stronger safety measures and proper human supervision. Liguori said Google's growing involvement with the US and Israeli militaries was the main reason behind her decision to resign. AI researcher Andreas Kirsch also resigned earlier in October. In an interview with the BBC, he said Google's systems for governance and oversight were “not sufficient or meaningful enough to deploy AI safely and responsibly in military and classified settings.”

Kirsch has also argued that current AI models are “not robust enough” for such uses. He raised concerns that Google appears to have little ability to track how its systems are used after they are handed over to the military. Both researchers have raised concerns about what happens once Google's AI models enter classified government systems. They fear that agreements allowing their use for “any lawful government purpose” leave the company with little practical control over how the technology is eventually used.

What Google says about AI safety Google has said employees have dedicated channels to raise ethical concerns and that its policies do not allow retaliation against those who speak up. The company has also previously said its military agreements include restrictions on mass surveillance and autonomous weapons. Their concerns centre on whether the company can ensure that its safety rules are followed once its AI technology is being used in classified military environments.

The latest resignations follow months of internal opposition to Google's military agreements. In April, around 600 employees signed a petition urging the company's leadership to stay away from classified military work. Research scientist Alex Turner, who worked on AI safety at Google DeepMind for more than two years, resigned in June. In a blog post published in July, he said senior management had insisted that Google would not sign such a deal, but the agreement went ahead.

“At that point, I couldn’t stay at Google in good conscience, so I left,” Turner wrote. The resignations come as the Pentagon seeks to bring advanced AI technology into classified military networks.