Funding

Microsoft's AI Bet Pays Off: $3.2B Gain on Anthropic in One Quarter Nearly Matches OpenAI's Full-Year Return

Microsoft disclosed a $3.2 billion gain on its investment in AI lab Anthropic in Q4 fiscal 2026, nearly matching the $5 billion full-year gain from OpenAI. The gain boosted diluted EPS by $0.33, while OpenAI was marked down $600 million in the same quarter. Microsoft's dual investments in competing AI labs highlight its hedging strategy amid rapid AI market evolution.

Neura News

Neura News

Neura Market Editorial

July 29, 20263 min read
Microsoft's AI Bet Pays Off: $3.2B Gain on Anthropic in One Quarter Nearly Matches OpenAI's Full-Year Return

Microsoft disclosed a $3.2 billion gain on its investment in AI lab Anthropic during the fourth quarter of its fiscal year 2026, a figure that nearly matched the $5 billion gain it recorded from OpenAI over the entire fiscal year. The revelation came in the company's earnings report for the period ended June 30, 2026, published on July 29, 2026.

A $3.2B Quarterly Boost from Anthropic

The $3.2 billion gain on Microsoft's Anthropic investment boosted diluted earnings per share by $0.33. Microsoft reported diluted EPS of $4.81 for the fourth quarter. The gain stems from Microsoft's $5 billion investment in Anthropic in November 2025, which was part of a circular agreement under which Anthropic committed to buying $30 billion worth of Azure services. Microsoft does not routinely update the value of its Anthropic investment each quarter, but the company disclosed the gain because it was noteworthy.

OpenAI Investment: A Mixed Quarter, Strong Year

Microsoft's investment in OpenAI, the other major AI lab in its portfolio, did not fare nearly as well in the fourth quarter. The company marked down its OpenAI investment by about $600 million in Q4, reducing diluted EPS by about $0.07. The write-down was mostly a rounding error for Microsoft, given its massive scale. However, the OpenAI investment looks much better when viewed on a full-year basis. For fiscal year 2026, Microsoft's OpenAI investment generated a $5 billion gain and added $0.67 to EPS. Microsoft owns about 27% of OpenAI and also receives revenue-share payments from the AI lab, though it does not report the amount.

The #1 Newsletter in AI

Stay ahead of the AI curve

The most important updates, news, and content — delivered weekly.

No spam. Unsubscribe anytime.

Microsoft's Overall Financial Performance

Microsoft reported Q4 revenue of $90 billion and net income of $35.8 billion. For the full fiscal year 2026, the company posted revenue of $331.8 billion and net income of $133.7 billion. Microsoft reported FY2026 diluted EPS of $17.95. The $3.2 billion gain on Anthropic in one quarter was nearly as much as the $5 billion gain on OpenAI for the entire fiscal year, highlighting the volatility and potential of Microsoft's dual bets on competing AI labs.

The Strategic Rationale Behind Dual AI Investments

Microsoft's investments in both Anthropic and OpenAI reflect its strategy to hedge its bets across the rapidly evolving AI landscape. The $5 billion investment in Anthropic in November 2025 came with a $30 billion Azure services commitment, ensuring that Anthropic's cloud computing needs are met by Microsoft's infrastructure. Meanwhile, Microsoft's 27% stake in OpenAI gives it a significant equity position in the creator of ChatGPT. The company discusses its OpenAI investment quarterly, but it does not routinely update the value of its Anthropic investment, making the $3.2 billion gain a notable disclosure.

Related on Neura Market

More from Neura News

Research

World Rehearsal Lets LLM Agents Learn Without an Environment

A new training method called EnvACE lets large language model agents practice inside their own heads instead of relying on external environments. The approach, described in a paper submitted to arXiv on 6 Aug 2026, replaces real-world interaction during training with what the authors call world rehearsal. The result is a policy that internalizes how actions change the world, and it performs strongly across four benchmarks.

Aug 7·4 min read
Research

New Benchmark Measures How Multi-Agent Systems Fail and Recover

OrchestraBench, a new benchmark introduced in an arXiv paper, uses controlled failure injection to measure how multi-agent systems fail and recover. It introduces metrics like cascade radius and per-failure-mode recovery, revealing that simple routers fail on adversarial cases while intent-reasoning models succeed. The benchmark also identifies three tiers of failure handling and shows that blind retry amplifies latent faults.

Aug 7·4 min read