Salesforce CEO Marc Benioff's prediction about Microsoft and OpenAI's partnership may have just manifested — and it's not a pretty look for the ChatGPT maker

There seems to be trouble brewing that could potentially drift Microsoft and OpenAI further apart. According to The Information, Microsoft is reportedly developing in-house reasoning models to give OpenAI a run for its money.

The report details that the software giant has been testing models from DeepSeek, Meta, and Elon Musk's xAI in Copilot, suggesting that the company might have further plans to sever its ties with OpenAI and its overreliance and dependence on ChatGPT.

This new development doesn't come as a surprise. Last year, a separate report detailed Microsoft's complaints about OpenAI's technology across its tech stack.

It suggested the Redmond giant could be moving away from OpenAI's AI products, the GPT-4 model, because it's too expensive and isn't fast enough to meet its enterprise customers' requirements.

Microsoft and OpenAI share a complicated partnership. As you may know, the Redmond giant made a multi-billion dollar investment in the ChatGPT maker, making it OpenAI's exclusive cloud partner for its AI advances.

However, this clause changed shortly after OpenAI unveiled its $500 billion Stargate project, designed to facilitate the construction of data centers across the United States. Microsoft holds the “right of first refusal" and remains the first option to host OpenAI workloads in its cloud infrastructure and services.

It's worth noting, OpenAI can outsource the services from other companies if Microsoft is unable to meet the requirements. Salesforce CEO Marc Benioff predicted Microsoft won't use OpenAI's technology across its tech stack, including Copilot, in the future because it's too expensive and isn't fast enough.

Microsoft has been placed under fire for its recent Copilot update, which has seemingly degraded the service's user experience. Microsoft staffers have blatantly branded the upgrade "a step backward."

As you may know, Microsoft's partnership with OpenAI dictates that the tech giant retains the right to use the AI firm's intellectual property. However, OpenAI is seemingly remaining tight-lipped about how it developed and trained its o1 reasoning model.

Microsoft AI CEO Mustafa Suleyman previously requested OpenAI employees to provide documentation about o1's development. However, his request was blatantly turned down.

Salesforce's CEO claims, "Mustafa Suleyman and Sam Altman aren't best friends." He attributed his deductions to last year's Davos conference, where both executives were on the panel. He claimed that they seemed uneasy around each other, potentially highlighting friction between Microsoft and OpenAI.

The once-considered "best tech bromance" has seemingly turned into a "winner takes it all" competition between Microsoft and OpenAI. Microsoft is reportedly planning to sell access to its in-house reasoning model (dubbed MAI) to developers, taking on OpenAI dominance in the AI landscape.

OpenAI's AI strategies are unrealistically expensive

Over the past few months, multiple reports have emerged indicating that OpenAI was on the verge of bankruptcy, with projections of making a $5 billion loss within 12 months.

However, the ChatGPT maker survived the bankruptcy claims after key stakeholders, including Microsoft, NVIDIA, Thrive Capital, and SoftBank raised $6.6 billion through a round of funding, pushing its market cap well beyond $157 billion.

Interestingly, market analysts and experts predict that OpenAI might still be in trouble, predicting that Microsoft could acquire the AI firm within 3 years.

A recent report highlighted OpenAI's potential plans to hike the cost of its next-gen AI models with PhD-level reasoning and thinking to $20,000 per month.

The report brewed controversy among the community, with some indicating that they'd rather wait for DeepSeek to distill the model and provide a similar iteration for free. Perhaps the highlighted changes in OpenAI's model pricing could be a bold attempt to meet the exorbitant cost of developing and training AI models.

This is especially true, as investors are mounting pressure on OpenAI to evolve into a for-profit entity, following the latest round of funding. This will help the company keep outsider interference and hostile takeovers at bay.

As you know, OpenAI's attempt to transition has received backlash from multiple parties, including Elon Musk (who helped co-found the firm). The billionaire has already filed two lawsuits against the firm, citing a stark betrayal of its founding mission and alleged involvement in racketeering activities.

While Musk's bid to block OpenAI's evolution into a for-profit entity was recently denied, other aspects detailed in his suit can proceed to trial.

Elon Musk recently offered to buy OpenAI for $97.4 billion. However, OpenAI CEO Sam Altman quickly turned down the offer, indicating the company and its mission aren't for sale.