Found an error?
Report now
More than six years ago, Google pledged not to make artificial intelligence available for weapons systems. In a new version of the AI principles, this is no longer the case. (archive image)
Keystone
Google has abandoned its promise not to develop artificial intelligence for weapons. The internet giant has updated the principles for its AI activities.
The new version no longer includes the voluntary commitment made in 2018. The reason given for the new version in a blog post was that technology has changed rapidly since 2018. The entry did not specifically address the resolutions that are no longer mentioned. The new principles generally state that the aim is to bring AI into line with "widely accepted principles of international law and human rights".
In contrast, it was mentioned separately in 2018 that Google would not develop AI that could be used for surveillance that violates internationally recognized standards. It would also not work on technologies that could be generally harmful.
The 2018 principles were adopted following internal opposition within the company to its involvement in a US military drone program.
05.02.2025, 01:33
