Google closes AI deal with Pentagon

Contents

Google is apparently entering into the controversial AI deal with the US Department of Defense, despite the resistance of over 600 of its AI experts. This is reported by the news portal The Information, citing sources familiar with the matter. This allows the Pentagon to use Google's AI models for “any lawful, governmental purpose” in the future, including classified military tasks.

By signing the contract, Google becomes the next of the leading US AI companies to fully provide its AI models to the Pentagon. In recent months, OpenAI and xAI have already entered into similar contracts with the Pentagon. Their leading AI models can now be operated on classified intelligence networks.

Internal criticism without effect

Previously, über 600 leading AI experts from Google had protested against the impending contract signing. In a letter, they urged CEO Sundar Pichai to reject the use of Google's AI models for classified military purposes. After all, Google cannot prevent the misuse of its AI technology if it is used secretly, the AI experts explained, including many from the research laboratory DeepMind AI.

“We want AI to benefit humanity, not to be used in inhumane or extremely harmful ways,” write the concerned employees. “This includes lethal autonomous weapons and mass surveillance, but goes beyond that.”

Google wants to support national security

According to The Information, the contract between Google and the Pentagon does include a clause that excludes the use of Google's AI systems for domestic mass surveillance and for autonomous weapon systems without human supervision. However, the agreement does not grant Google any rights to control or even object to the government's operational decisions. In effect, the unrestricted use of Google's AI models for military purposes would thus be enabled.

Google told the news agency Reuters that it supports government agencies in both classified and unclassified operations. However, the company continues to commit to not using AI technology for domestic surveillance or autonomous weapons.

“We believe that providing API access to our commercial models, including Google infrastructure, in compliance with industry-standard practices and terms, represents a responsible approach to supporting national security,” a Google spokesperson told Reuters.

Pentagon: AI allegedly only for “lawful purposes”

According to the Reuters report, the Pentagon stated that it has no interest in using AI systems for mass surveillance or for human-independent weapon systems. It only wants to use AI for “lawful purposes.” However, what the ministry classifies as legal is changing. In the recent past, for example, the USA repeatedly attacked civilian boats off the coast of Venezuela. The boats were allegedly smuggling drugs.

Furthermore, AI technology has reportedly been used multiple times in recent months for US military operations. During the apprehension of former Venezuelan leader Nicolás Maduro, AI is said to have supported the operation. And in the Iran conflict, there are increasing indications that Palantir's AI system, Maven Smart System (MSS), has already been used to select targets for military strikes. An incorrect assessment by the AI system could have been the reason for the fatal US airstrike on an Iranian school in Minab on February 28. (rah)