Google recently changed its artificial intelligence (AI) guidelines. It no longer bans the use of AI for making weapons or surveillance tools.

Experts think AI could soon be common in military use, but there are worries, especially about weapons that can kill without human input. AI can be very useful in war, changing everything from office work to combat. But there’s debate about how to control this powerful tech, especially in war where AI could decide to kill many people. The Doomsday Clock, which warns about global threats, also pointed out dangers from AI in military use.

Human Rights Watch finds this change very worrying, BBC News reports. They say AI can make it hard to tell who is responsible for decisions on the battlefield, where lives are at stake. They argue that we need laws to control AI, not just voluntary guidelines.

Originally, Google’s founders had a motto, “don’t be evil,” but then the company changed it to “do the right thing.” Google employees have protested against using AI for harmful military projects before.

Google’s James Manyika and Demis Hassabis have written a blog post defending this shift, saying businesses and governments should cooperate on AI to help with national security. They argue that AI needs updating to match new technology while still holding onto values like freedom and human rights. They believe democracies should guide AI development.

Toward innovation and AGI

Google updated its Frontier Safety Framework to handle new risks from advanced AI models, focusing on security, misuse prevention, and deceptive AI behavior. AI principles have evolved globally, with frameworks from the G7 and others guiding AI use. Google updated its AI Principles to focus on bold innovation, responsible deployment, and collaborative progress. They aim to develop AI that respects human rights and international laws, while considering AI’s potential in various fields.

As AI evolves, Google promises to adapt its approach, especially with artificial general intelligence (AGI), emphasizing both innovation and safety.

Let us know your thoughts! Sign up for a Mindplex account now, join our Telegram, or follow us on Twitter.