Sam Altman, left, has pushed Microsoft executives like Kevin Scott, right, to give his company more computing power.
Credit...Grant Hindsley for The New York Times

Microsoft and OpenAI’s Close Partnership Shows Signs of Fraying

The “best bromance in tech” has had a reality check as OpenAI has tried to change its deal with Microsoft and the software maker has tried to hedge its bet on the start-up.

by · NY Times

Last fall, Sam Altman, OpenAI’s chief executive, asked his counterpart at Microsoft, Satya Nadella, if the tech giant would invest billions of dollars in the start-up.

Microsoft had already pumped $13 billion into OpenAI, and Mr. Nadella was initially willing to keep the cash spigot flowing. But after OpenAI’s board of directors briefly ousted Mr. Altman last November, Mr. Nadella and Microsoft reconsidered, according to four people familiar with the talks who spoke on the condition of anonymity.

Over the next few months, Microsoft wouldn’t budge as OpenAI, which expects to lose $5 billion this year, continued to ask for more money and more computing power to build and run its A.I. systems.

Mr. Altman once called OpenAI’s partnership with Microsoft “the best bromance in tech,” but ties between the companies have started to fray. Financial pressure on OpenAI, concern about its stability and disagreements between employees of the two companies have strained their five-year partnership, according to interviews with 19 people familiar with the relationship between the companies.

That tension demonstrates a key challenge for A.I. start-ups: They are dependent on the world’s tech giants for money and computing power because those big companies control the massive cloud computing systems the small outfits need to develop A.I.

No pairing displays this dynamic better than Microsoft and OpenAI, the maker of the ChatGPT chatbot. When OpenAI got its giant investment from Microsoft, it agreed to an exclusive deal to buy computing power from Microsoft and work closely with the tech giant on new A.I.

“We’re deeply grateful for our partnership with Microsoft; the early big bet they took on us and the vast compute resources they’ve provided have been essential to our research breakthroughs, benefiting both companies greatly,” Mr. Altman said in a statement Thursday. “We are excited and committed to pursuing our shared vision and achieving even greater things together far into the future.”

Over the last year, OpenAI has been trying to renegotiate the deal to help it secure more computing power and reduce crushing expenses while Microsoft executives have grown concerned that their A.I. work is too dependent on OpenAI. Mr. Nadella has said privately that Mr. Altman’s firing in November shocked and concerned him, according to five people with knowledge of his comments.

Since then, Microsoft has started to hedge its bet on OpenAI.

“We have continued to invest in OpenAI at many discrete points in the partnership,” Mr. Scott, Microsoft’s chief technology officer, said in a recent interview. “We are certainly the very largest investor of capital in them.”

But in March, Microsoft paid at least $650 million to hire most of the staff from Inflection, an OpenAI competitor. Inflection’s former chief executive and co-founder, Mustafa Suleyman, oversees a new Microsoft group that is working to build A.I. technologies for consumers based on OpenAI software. He is also the point person for Microsoft’s long-term effort to build technologies that could replace what the company is getting from OpenAI, according to two people familiar with Microsoft’s plans.

“Microsoft could be left behind if it is only using OpenAI technologies,” said Gil Luria, an analyst at the investment bank D.A. Davidson. “It is a real race — and OpenAI may not win it.”

Some OpenAI executives and employees, including Mr. Altman, are angered that Mr. Suleyman is at Microsoft, according to five people familiar with the relationship between the two companies. Mr. Suleyman’s team is part of a group of Microsoft engineers who work directly with employees at OpenAI. Dozens of Microsoft engineers work on-site at OpenAI’s offices in San Francisco and use laptops provided by OpenAI that are set up to maintain the start-up’s security protocols.

Some OpenAI staff recently complained that Mr. Suleyman yelled at an OpenAI employee during a recent video call because he thought the start-up was not delivering new technology to Microsoft as quickly as it should, according to two people familiar with the call. Others took umbrage after Microsoft’s engineers downloaded important OpenAI software without following the protocols the two companies had agreed on, the people said.

(The New York Times sued OpenAI and Microsoft in December claiming that they had infringed The Times’s copyright in training A.I. systems.)

After Microsoft backed away from the discussions about additional funding, OpenAI was in a bind. It needed more cash to keep its operations going, and its executives chafed at the exclusivity of the contract. Over the past year, the A.I. company repeatedly tried to negotiate a lower cost and allow it to buy computing power from other companies, according to seven people familiar with the discussions.

In June, Microsoft agreed to an exception in the contract, six people with knowledge of the change said. That allowed OpenAI to sign a roughly $10 billion computing deal with Oracle for additional computing resources, according to two people familiar with the deal. Oracle is providing computers packed with chips suited to building A.I., while Microsoft provides the software that drives the hardware.

And in recent weeks, OpenAI and Microsoft negotiated a change to a future contact that reduces how much Microsoft will charge the smaller company for computing power, although the exact terms were unclear, according to a person familiar with the change.

While it was looking for computer power alternatives, OpenAI also raced to broaden its investors, according to two people familiar with the company’s plan.

Part of the plan was to secure strategic investments from organizations that could bolster OpenAI’s prospects in ways beyond throwing around money. Those organizations included Apple, the chipmaker Nvidia, and MGX, a tech investment firm controlled by the United Arab Emirates.

Mr. Altman and OpenAI had been discussing potential partnerships with Apple for years. In 2022, as OpenAI was developing the technologies that would drive ChatGPT, Mr. Altman and Kevin Scott, Microsoft’s chief technology officer, met with executives at Apple to explore ways the three companies might work together, according to two people familiar with the meeting. That meeting eventually led Apple to agree to put ChatGPT on the iPhone earlier this year.

Nvidia was an important partner because it designed the computer chips that OpenAI needed to build its A.I. technologies. MGX was part of an ambitious OpenAI effort to build new computer data centers across the globe.

Earlier this month, OpenAI closed a $6.6 billion funding round led by Thrive Capital, with additional participation from Nvidia, MGX and others. Apple did not invest, but Microsoft also participated in the funding round.

OpenAI expected to spend at least $5.4 billion in computing costs through the end of 2024, according to documents reviewed by The New York Times. That amount was expected to skyrocket over the next five years as OpenAI expanded, soaring to an estimated $37.5 billion in annual computing costs by 2029, the documents showed.

It is not clear how much the recent tweaks to the partnership between OpenAI and Microsoft will alter that trajectory, but Microsoft executives were happy with the changes, according to a person familiar with the company’s strategy. The tech giant can continue to benefit from OpenAI’s improving technologies, while the start-up continues to pay the tech giant for substantial amounts of computing power.

Still, OpenAI employees complain that Microsoft is not providing enough computing power, according to three people familiar with the relationship. And some have complained that if another company beat it to the creation of A.I. that matches the human brain, Microsoft will be to blame because it hasn’t given OpenAI the computing power it needs, according to two people familiar with the complaints.

Oddly, that could be the key to getting out from under its contract with Microsoft. The contract contains a clause that says that if OpenAI builds artificial general intelligence, or A.G.I. — roughly speaking, a machine that matches the power of the human brain — Microsoft loses access to OpenAI’s technologies.

The clause was meant to ensure that a company like Microsoft did not misuse this machine of the future, but today, OpenAI executives see it as a path to a better contract, according to a person familiar with the company’s negotiations. Under the terms of the contract, the OpenAI board could decide when A.G.I. has arrived.

At an A.I. conference in Seattle this month, Microsoft didn’t spend much time discussing OpenAI. Asha Sharma, an executive working on Microsoft’s A.I. products, emphasized the independence and variety of the tech giant’s offerings.

“We definitely believe in offering choice,” Ms. Sharma said.

Karen Weise and Tripp Mickle contributed reporting.

Explore Our Coverage of Artificial Intelligence

News and Analysis

  • Cerebras, a chip company with bold ambitions to take on Nvidia, filed for an I.P.O., taking a key step toward being among the first A.I. companies to go public since the release of ChatGPT.
  • Gov. Gavin Newsom vetoed a California A.I. safety bill, blocking the most ambitious proposal in the nation aimed at curtailing the growth of the technology.
  • At an event in Silicon Valley, Meta exhibited a range of products, including new smart glasses, meant to blend the real world and virtual reality with a healthy dose of A.I.

The Age of A.I.