Google is making an interesting change to AI policy – Aroged

Since the launch of Chatgpt in 2022, companies have increasingly paying more attention to artificial intelligence. It is important to do a strong policy, in which it is explained what can and may (and what not) with AI. Companies that develop AI applications also pursue such a policy. Google, one of the most important AI developers of the moment, makes a substantial adjustment to the policy. The company scraps the clause in which it says that it is not to complicate AI applications if it is likely that they cause damage.

Demis Hassabis, CEO of Google-daughter company DeepMind, explains together with James Manyika, SVP for Research, Labs, Technology & Society, the change in a blog post. The CEO begins to frame the evolution of AI. In 2018 it was still a technology that was only used by researchers, but in 2025 the technology penetrated into daily life, such as smartphones and the internet that previously did.

Shared guidelines

Shared guidelines, given the omnipresence of technology, are extremely important. Just that fact requires Google to revise the AI ​​policy and to cancel the safety clause. Earlier, the company promised not to make AI applications that can be used as a weapon or surveillance. The company no longer imposes that obligation, although that does not mean that it has free play. Hassabis and Manyika refer to the rules De G7, ISO and Democratic states around the world. The company will use those shared standards, instead of its own set of rules when it comes to the safety of AI applications.

That does not mean that Google will let all the principles sail: the company’s guidelines can still be consulted on the Webste AI.Google. These reworked guidelines focus on three core values: daring innovation, responsible development and contribution and collaborative progress. With those values ​​in mind, Google wants to continue to build AI applications in the coming years, which should improve the lives of users.