Compiled by Futu News
According to informed sources, Google abruptly withdrew from the challenge after submitting a successful proposal in a $100 million Pentagon competition. The competition aimed to develop autonomous drone swarm technology for voice control.
Another person familiar with the matter revealed that the company notified the government on February 11 that it would no longer participate in this program, which aims to create the technology required for controlling drone swarms, just weeks after submitting its proposal. Relevant records reviewed by Bloomberg show that this decision was made following an internal ethics review. According to these records, Alphabet's Google cited 'insufficient resources' as the official reason for withdrawing from the competition.
Records indicate that several Google employees involved in this project expressed disappointment over the decision to withdraw from the competition. It remains unclear how widely known within the company Google’s initial participation in this autonomous drone technology challenge was. However, hundreds of AI researchers at Google have broadly opposed the use of the company's cutting-edge technology for classified military work.
Due to the sensitivity of the matter, these individuals requested anonymity.
Google's participation in and subsequent withdrawal from the competition had not been previously reported.
After evaluating the project, we decided not to submit a bid so that we could focus on initiatives where our models can have the greatest impact,' a spokesperson for Google's public sector told Bloomberg. The spokesperson stated that the company evaluates hundreds of government opportunities annually and prioritizes bidding on projects that best align with current resource and technological strengths. The spokesperson did not respond to questions about the internal ethics review.
This Pentagon initiative, jointly led by the Defense Autonomous Warfare Group under the Special Operations Command and the Defense Innovation Unit, envisions commanders being able to direct drone swarms by converting voice commands like 'turn left' into digital instructions.
The U.S. Special Operations Command referred related inquiries to the Pentagon, which declined to comment. The Defense Innovation Unit did not respond to requests for comment.
Companies such as OpenAI, Palantir, and xAI have been selected to participate in this competition, which is scheduled to unfold in multiple phases over six months. Subsequent phases of the competition require the development of 'target-related situational awareness and sharing' and capabilities spanning 'from launch to termination.'
Anthropic PBC also applied to participate in the competition but was not selected. Although CEO Dario Amodei has reservations about integrating large language models into fully autonomous weapons, the company assessed that its submission did not violate its established red lines and would contribute to the development and testing of this controversial new technology.
Google's decision to withdraw from this six-month challenge comes as leading AI companies and their employees grapple with the implications of assisting the U.S. in developing autonomous lethal weapon systems. The role of technology in enabling potentially high-risk new weapons has been a recurring point of contention between the Pentagon and Silicon Valley.
According to the organizers of the event, on Monday, hundreds of Google AI researchers signed a letter to Alphabet CEO Sundar Pichai, urging him to reject providing the company's AI systems for classified workloads related to U.S. defense missions. This followed a report by The Information that Google and the Pentagon had signed a new AI agreement allowing AI to be used for 'any lawful government purpose,' and that Google would have no veto power over legitimate government operational decisions, including classified work.
A Google spokesperson told Bloomberg that the company had revised its existing contract with the Pentagon, providing access to its AI models without involving the development of custom models for the Pentagon, which the spokesperson described as 'a responsible way to support national security.'
The spokesperson added: 'We remain committed to the consensus between the public and private sectors that AI should not be used for domestic mass surveillance or autonomous weapons without appropriate human oversight.'
The Pentagon declined to comment on whether it had reached an agreement with Google to revise the existing deal or on the related terms.
Editor/Rocky