C3.ai, Inc. (AI) stock intrinsic value fair value estimate & owner earnings analysis

C3.ai sells enterprise AI software (the C3 Agentic AI Platform and applications for asset performance, supply chain, procurement and similar uses) to large organizations in energy, manufacturing, healthcare and government. Revenue fell 36% in the fiscal year ended April 30, 2026, from $389.1 million to $250.3 million. The company attributes this to disruption from a reorganization of its sales and services teams, health limitations of its founder and CEO, and weak execution. Thomas Siebel returned as CEO on May 8, 2026. On a per-share owner-earnings basis (operating cash flow, less working-capital swings, less stock-based compensation, less maintenance capital spending), the company has lost money in every year we can measure. Owner earnings per share were about -$1.67 in fiscal 2022, -$2.85 in 2023, -$2.47 in 2024, -$2.15 in 2025 and -$3.32 in 2026. The five-year median is about -$2.47. Growth in owner earnings in the latest year was negative (the loss widened by about 55%). The main reason is that stock-based compensation, $263.7 million in fiscal 2026, is larger than the entire year's revenue, while operating cash flow was -$190.2 million. We treat stock compensation as a real cost paid to employees in shares, so it is deducted. Bear, base and bull cases are each valued at $0 per share. In each case the question is what evidenced, already-secured earning power there is to discount. Contracted revenue is real: $203.1 million of remaining performance obligations, $129.3 million of it expected within twelve months. But it is smaller than the $235.1 million a year earlier, and the company's costs run well above it. A restructuring plan approved on February 24, 2026 aims to cut headcount and non-employee costs, with implementation expected to finish by the second quarter of fiscal 2027, and management says fiscal 2027 operating expenses should fall. That is a plan and a target, not a demonstrated result, and stock compensation has risen every year. We therefore give no credit for a turnaround, for renewed growth, or for margin improvement; no earning stream we can name and evidence remains to value. The bear case is worse than the base only in that cash keeps draining: cash and marketable securities fell from $742.7 million to $575.4 million in fiscal 2026. The company has no meaningful debt that would push any case below zero. This method values the earning power of the business and does not add the balance-sheet cash to that value. Shareholders own that $575.4 million, so the shares are not worthless if the company burns it more slowly than the past year's pace, and a return to positive owner earnings after the restructuring would change our view quickly. Also, a single customer-concentration risk is disclosed: a limited number of customers have historically accounted for a substantial share of revenue, and some contracts can be cancelled without a significant penalty. What would make this wrong: the restructuring delivers cost cuts that bring operating cash flow above stock compensation while revenue stabilises around the contracted base. That would warrant a fresh estimate after several quarters of evidence. No margin-of-safety price is set until that condition is met. Note: the founder controls about 49.6% of the voting power through Class B shares as of June 10, 2026, and there are securities class-action suits against the company and its officers.