Investors sue Microsoft over Copilot claims as GitHub seeks cloud capacity

Microsoft has entered a difficult stretch for its AI business as investors challenge its Copilot disclosures and GitHub seeks more infrastructure capacity for AI-heavy developer workflows.

The City of St. Clair Shores Police and Fire Retirement System filed a class action in U.S. District Court in Seattle against Microsoft and several executives, including CEO Satya Nadella. The complaint says Microsoft gave investors false or misleading statements about Copilot adoption, Microsoft 365 conversion rates, and the strength of its in-house AI models.

The case centers on Microsoft's public statements about Copilot, the AI assistant line that spans Microsoft 365, Windows, security tools, developer products, and business software. Investors say Microsoft did not give enough detail about weak customer uptake and model performance before the company reported slower Azure growth Jan. 28.

Microsoft said it knows about the complaint, rejects the claims, and plans to defend its public statements in court.

The filing says Microsoft had reached 15 million paid Microsoft 365 Copilot seats out of 450 million Microsoft 365 users. After Microsoft released those fiscal second-quarter results, its share price fell more than $48 per share, according to the complaint.

For compliance and investor relations teams, the lawsuit points to a direct risk in AI marketing. Companies that tell investors about AI adoption need numbers that match internal sales data, renewal trends, churn, and product usage. Broad claims about demand can create securities risk when executives know customers have not converted at scale.

GitHub faces a separate pressure point. Microsoft bought GitHub in 2018, and the developer platform has tried to shift more workloads to Azure. GitHub has also seen new strain from AI coding tools and agentic development, where automated systems generate code, run tasks, and call platform services at higher volume than conventional human workflows.

GitHub said its community has grown at a pace it has not seen before. The company said the spike in agentic development that began late last year tested its infrastructure limits. GitHub said it will speed up its Azure migration and explore a multicloud strategy to add capacity, compute elasticity, and scale.

Reports say GitHub may use Amazon Web Services capacity to support that demand. That choice would carry an awkward signal because Microsoft owns Azure, but many infrastructure teams use more than one cloud provider to reduce concentration risk, add regional capacity, or handle burst demand.

The two issues share one theme: Microsoft has tied much of its growth story to AI, and customers now need proof that the products and infrastructure can carry that growth. Copilot needs paid adoption inside Microsoft 365 accounts. GitHub needs capacity for AI coding workflows. Azure needs enough compute to serve both Microsoft products and cloud customers.

Compliance teams should treat the lawsuit as a reminder to test AI disclosures before executives use them in earnings calls, investor decks, or SEC filings. Legal, finance, product, and sales leaders should compare public claims with source data, including paid seats, active users, model benchmarks, customer retention, and competitive losses.

Cloud and platform teams should take a similar lesson from GitHub's capacity strain. AI workloads can change traffic patterns fast. Agentic tools can create more API calls, more jobs, and more background compute than teams planned for under older usage models. Capacity plans need stress tests that account for automated development agents, not only human users.

Microsoft can still defend the lawsuit and expand Copilot sales. GitHub can still add capacity and improve reliability. The near-term challenge sits in execution: Microsoft needs to prove that its AI claims, customer adoption, and cloud infrastructure can line up under investor, customer, and regulator scrutiny.