In the deleted section of the AI ethics principles, Google pledged not to develop or deploy AI for use in surveillance systems, weapons, or technologies designed to harm people.
Demis Hassabis, CEO of Google DeepMind, and James Manyaka, Senior Executive Director of Technology and Society at Google, shared new "core principles" in AI development. They include innovation, collaboration, and "responsible" AI development, but without giving specific details.
This clause of Google's policy put the company at a slight disadvantage relative to its competitors in the AI market. Meta's Llama and OpenAI's ChatGPT technologies are authorized for military use. Amazon has entered into an agreement with Palantir, a state-owned software manufacturer, and Anthropic sells its Claude AI to the US military and intelligence services.
Despite its commitment, Google has already worked on various military contracts, including Project Maven, a 2018 Pentagon project in which Google used artificial intelligence to analyze drone video, and the 2021 Project Nimbus military cloud contract with the Israeli government.