Google Removes Pledge to Abstain from Using AI for Harmful Applications and Weapons

Google has updated its public AI principles page and dropped its pledge to not use AI for harmful applications like weapons and surveillance. 

The said pledge used to sit in the “applications we will not pursue” section on their page and Bloomberg was the first one to point out that it’s no longer there. 

This change could be simply to be able to continue holding ground amid the rapidly increasing competition in the AI space, especially after China has entered the space with its allegedly revolutionary DeepSeek.

The new AI principles say that Google will work to mitigate unintended or harmful outcomes and will avoid unfair bias. While that sounds like a relief, it’s still uncertain what the agenda behind this change of policy is. 

This update primarily reflects Google’s growing ambitions to offer AI to more users, including governments. The new AI principles say that Google will “proceed where we believe that the overall likely benefits substantially exceed the foreseeable risks and downsides.”

It’s worth noting that OpenAI recently launched a special ChatGPT Enterprise variant for the US government. Is this AI companies looking to curry favor with Trump? Could be.

Google’s History With AI Ethics 

Google has aggressively pursued federal government contracts before the AI principles were established in 2018, when the tech giant refused to renew its contract with the government due to strong protests from the internal Google team.

A large number of employees had signed petitions opposing the contract and many resigned from the company due to its involvement in the project. The contract was called Project Maven and it basically helped the government analyze and interpret drone videos using artificial intelligence. 

Another contract, called Pentagon cloud, which was speculated to be worth $10 billion, was also turned down by the company saying it didn’t align with its values.

Just last year, a series of protests took place over Project Nimbus, a Google contract with Amazon services to provide cloud computing and AI services to the Israeli government and military. 

Google terminated more than 50 employees as a consequence of this and released a statement saying the contract didn’t violate any of the company’s AI principles. 

The fact of the matter, as revealed by The New York Times, was that just a few months before this happened, Google officials themselves were seen expressing concern about the contract harming the company’s reputation as its cloud services could be used for facilitation of human rights violations. 

Currently, the updated guidelines restrict the internal forum from a discussion about geopolitical content, international relations, military conflicts, economic actions, and territorial disputes. 

When CNBC asked for comment on the guidelines, Google did not immediately respond to a request for comment. 

Alpa is a tech writer and editor with a wealth of experience in alternative finance, fintech, cryptocurrency, app security software, and the medical industry. She’s passionate about breaking down complex topics and sharing informative content that provides value.

View all articles by Alpa Somaiya