Microsoft accused of trying to manipulate AI pricing via its OpenAI investment

Microsoft, which is no stranger to antitrust accusations, is being sued for allegedly manipulating AI prices via its investment in OpenAI.

The lawsuit, filed Monday in the US District court in the Northern District of California, argues that Microsoft has used its OpenAI influence to vastly increase AI prices.

“ChatGPT prices were inflated since the service’s inception, with price levels reaching an eye-popping 100 to 200 times competitors’ prices on a per-token basis amidst a February 2025 price war,” the filing said. “A secretive agreement struck between OpenAI and Microsoft early in OpenAI’s development allowed Microsoft to control the supply of compute to its horizontal competitor’s products. It used an exclusivity clause to restrict OpenAI’s product output, and to impose a price–or, conversely, output and quality–floor on its competitor OpenAI’s ChatGPT products.”

Contacted by Computerworld, Microsoft declined to address the lawsuit’s specific details, but did say in an email, “while we are still reviewing the details of the complaint, we believe that our OpenAI partnership promotes competition, innovation, and responsible AI development.”

Industry observers were skeptical that the litigation would have much of an impact, given the large number of major companies selling AI, including Google, Amazon, and Anthropic, that are presumably beyond Microsoft’s direct influence.

Taps into policy concerns

Abhishek Singh, a partner at the Everest Group, said that the litigation might encourage regulatory efforts, but that he doubts it would have any direct impact on Microsoft.

“It taps into a genuine policy concern about the concentration of power in AI infrastructure and pricing, but antitrust cases like this are hard to win,” Singh said. “The plaintiffs will have to show not just dominance, but deliberate collusion and measurable consumer harm, which is a high bar.”

But Singh added that even if this specific lawsuit doesn’t go anywhere, it could easily prompt secondary reactions from legislators and regulators.

“In my personal view, the case is not frivolous, but it doesn’t have strong legs as an anti-competitive lawsuit. What it will likely do, however, is push the conversation toward transparency in AI pricing and infrastructure access,” Singh said. “Much of what enterprises perceive as price control in AI is, in reality, a reflection of scarce GPU supply and the costs of training and running ever-larger models, rather than market manipulation.”

If the lawsuit clears the initial legal hurdles and is allowed to proceed, Singh said, “the most plausible outcome is increased regulatory and market pressure for transparency rather than anything structural. Microsoft will argue that its partnership with OpenAI accelerated innovation and democratized access to generative AI—and that narrative is credible.”

Singh also observed that the essence of the monopolistic and anti-trust accusations have been argued before, and Microsoft has typically won those arguments. 

“[Because] the lawsuit primarily hinges on the claim that Microsoft effectively controls OpenAI and has used that position to distort AI pricing, it’s important to note that  regulators in the UK and EU have already examined Microsoft’s relationship with OpenAI and declined to classify it as a merger giving Microsoft control. Also, more recently, the partnership itself has evolved to allow OpenAI to source non-Microsoft compute, which weakens the argument that Microsoft has a lock on AI infrastructure,” Singh said. 

Customers should rethink AI contracts

The most critical aspect is whether this litigation will lead to any reimbursement for enterprises, which seems unlikely. But rebates aside, Singh suggests that this lawsuit does suggest some changes for enterprise IT to consider. 

“For enterprise IT executives, the key takeaway is to recognize that AI pricing and infrastructure access are becoming strategic variables. This lawsuit reinforces the need to seek transparency in AI cost structures, diversify vendor dependencies, and build flexibility into commercial arrangements,” Singh said. “The broader signal is about vigilance, understanding where market influence is concentrating, and ensuring enterprise strategies remain adaptable as the AI ecosystem matures.”

Cybersecurity consultant Brian Levine, a former federal prosecutor who today serves as the executive director of FormerGov, a directory of former government and military specialists, argues that Microsoft has had decades of experience fending off various antitrust accusations. For enterprise IT executives, though, the actions mean that they must work with their general counsel to make sure that any Microsoft contracts acknowledge antitrust possibilities, and that the agreements include wording that will protect the enterprise should court rulings prove unfavorable to Microsoft.

“If I was an enterprise CIO, the only immediate action is, if we are considering working with either of these entities, to write in a provision that gives [the enterprise] room to renegotiate to the extent that this case reaches a judgment or they reach a settlement that may impact the contract,” Levine said, adding, “but that language may be implied anyway.”

Douglas Brush, a special master with the US federal courts, said that enterprise IT must rethink all AI contracts in light of these kinds of antitrust accusations.

“The best approach is to use short contracts with re-openers, transparent pricing with safeguards, multiple cloud options, and an economics model that prioritizes consumption. This allows [enterprises] to benefit from falling prices, protect themselves when they rise, and keep the business running regardless of any single vendor’s motives,” Brush advised. “Budgeting needs to treat AI like a commodity input, not a fixed software license — [cost of goods sold] versus [operating expense]. … Quarterly repricing and automatic rebases to current schedules are table stakes.”

The lawsuit does acknowledge that OpenAI eventually slashed token prices “by as much as 80 percent,” but argued that merely proved its point. 

The ultimate price reduction “makes clear” that Microsoft’s actions were “the but-for and proximate cause of the price inflation/supply and output restriction,” the lawsuit said, adding, “but the restraint—and Microsoft’s control over it—still remains, lingering as a sword of Damocles over OpenAI wielded by one of its principal competitors.”