As Anthropic clashed with the Department of Defense last month over the use of artificial intelligence in warfare, a top Google executive quietly met with Pentagon officials to pitch the company as the perfect defense partner.
On Feb. 26, Thomas Kurian, the chief executive of Google Cloud, sat down with Emil Michael, the Pentagon official in charge of selecting A.I. tools for the Defense Department. Google already had business with the Pentagon, and Mr. Kurian offered a reliable, expanded supply of some of the best A.I. tools on the market, without all the noise, said two people with knowledge of the meeting, who spoke on the condition of anonymity to discuss private conversations.
By the time the dust settled, the Pentagon had cut off Anthropic and signed a deal with OpenAI. OpenAI faced backlash over the agreement, including from its own employees, and amended the terms days later. The Defense Department had also agreed to expand its use of Google’s technology, adding the company’s autonomous systems known as A.I. agents to its unclassified networks. The deal was publicly announced last week.
As Anthropic and OpenAI grapple with the political fallout of their work with the Pentagon, Google has been sitting pretty.
Unlike its start-up competitors, Google turns a significant profit. Its parent company, Alphabet, hauled in $34.5 billion in profit last quarter. The tech giant makes its own chips, runs its own cloud computing service and owns its own data centers, the massive server farms that power the cloud and A.I.
The company, which is synonymous with internet search, sits on an enormous trove of data to train its chatbot, Gemini, which is catching up with Anthropic’s Claude and OpenAI’s ChatGPT in tests that measure A.I. capabilities. While competing A.I. companies have been plagued by co-founder infighting, executive ousters and employee uprisings, Google has crept ahead by offering a relatively reliable and well-funded chatbot.
Google, which had sworn off military work in 2018, is now rebuilding its relationship with the Defense Department. And it is poised to benefit as it sidesteps its rivals’ controversies.
“The bets they made as a company to embrace machine learning a decade ago were really visionary,” said Michael Nathanson, a senior research analyst at the equity research firm MoffettNathanson, referring to the technology that underlies many modern A.I. systems. “In the past year, you’ve started to see some of their real advantages starting to flex.”
Google, OpenAI and Anthropic declined to comment. (The New York Times has sued OpenAI and Microsoft for copyright infringement of news content. Both companies have denied those claims.)
One of Google’s key advantages in the A.I. race is that it has heaps of money. Its bread-and-butter businesses like search, cloud and YouTube rake in cash. Alphabet brought in sales of nearly $114 billion in the most recent quarter, which ended in December, an 18 percent increase from a year earlier. And while the company said in February that it would double its spending to at least $175 billion this year to build more data centers, investors haven’t been spooked.
Google already has quite a few data centers, as well as experience in building and running them. The company also designs its own chips to power the technology, which it began using internally in 2015. That is saving Google time and money as rivals scramble to build data centers and spend billions on A.I. chips from Nvidia.
Google’s other businesses are also benefiting from A.I. Customers are signing up in droves to run compute-heavy projects on Google’s cloud infrastructure. Google’s cloud revenue rose 48 percent to $17.7 billion in the most recent quarter.
Last year, Google released Gemini 3.0, one of the top-performing A.I. models overall, coming in behind Claude, according to Vals AI, a company that tracks the performance of A.I. technologies.
While Gemini lags behind others when tested for some specific uses, like getting A.I. to write computer code, the company charges less than its competitors, said Rayan Krishnan, the chief executive of Vals AI.
“In a lot of ways they are lagging, but showing they can be competitive with time,” Mr. Krishnan said.
The A.I. improvement and cost savings have made Google attractive to big organizations, like the Pentagon, Mr. Krishnan said.
But there is still reluctance among some Pentagon officials to rely on Google, two people familiar with discussions between the Defense Department and Google said. That’s because the company dropped a military contract in 2018 in response to protests from employees who argued that A.I. should not be used in weapons, the two people added.
Several top A.I. researchers at Google, as well as at OpenAI, also recently signed a legal briefing to support two lawsuits that Anthropic filed against the Defense Department. Anthropic is challenging the Pentagon for designating it a supply chain risk after it clashed with the department over how to use A.I. in warfare.
The participation of Google employees in the legal briefing added to some officials’ concerns about the company, reminding decision makers in the Pentagon and elsewhere of the past protests against the use of A.I. in weapons, one former official with knowledge of the discussions said. Some Trump administration officials are worried that even if Google agrees to have its A.I. used widely, the company could bow out again, the former official added.
The Google and OpenAI employees who signed the filing did so in their personal capacities, said Nicole Schneidman, a lawyer who represents the A.I. experts. Still, tech workers who build A.I. feel empowered to push back on their companies. “These are people who are not afraid to walk if a company is not moving forward in a way that aligns with their core values,” she said.
Kate Conger is a technology reporter based in San Francisco. She can be reached at [email protected].
The post Google Sits Pretty as A.I. Rivals Compete for Pentagon Favor appeared first on New York Times.
