C3.ai, Inc.

source: coverage-next-full ticker: AI company: C3.ai, Inc. step: 01 title: Business Model Overview created: 2026-06-11

Step 01 — Business Model Overview: C3.ai, Inc. (AI)

1. Executive Summary

C3.ai is a pure-play enterprise AI platform company selling pre-built AI applications and a low-code/no-code AI development platform to large enterprises and government agencies. The company's defining characteristic is extreme vertical specificity — rather than selling horizontal AI tooling, it sells pre-built AI applications for oil-and-gas optimization, financial services compliance, defense/intelligence analytics, and manufacturing predictive maintenance. The business model has undergone a radical transformation over FY2023–FY2026: from a large-deal, direct-sales, alliance-dependent model toward a partner-led, consumption-based model with lower ACV but higher pipeline volume. [S1]

2. Value-Chain Layer Map

Layer                        C3.ai's Position
─────────────────────────    ──────────────────────────────────────────
Infrastructure (GPU/IaaS)    NOT HERE — relies on Azure/AWS/GCP
Foundation Models (LLM)      NOT HERE — uses OpenAI, Anthropic, others
AI Dev Platform (MLOps)      ✓ C3 Agentic AI Platform (low-code AI builder)
Pre-Built AI Applications    ✓ C3 AI Applications (300+ apps, 8 verticals)
Generative AI Layer          ✓ C3 Generative AI (enterprise search, agents)
Systems Integration          Partial — Professional Services (16% revenue)
End Customer                 Large enterprises (Fortune 1000) + Federal/DoD

Key insight: C3.ai occupies the application and platform layers, deliberately avoiding the commodity infrastructure race. Its moat thesis rests on pre-built, production-ready AI apps with embedded domain expertise (e.g., oil & gas yield optimization, anti-money-laundering detection) that would take years to replicate in-house. [S1]

3. Product Architecture

C3 Agentic AI Platform

Low-code/no-code environment for enterprises to build, deploy, and manage AI applications. Includes data integration, model management, workflow orchestration, and monitoring tools. Launched "Agentic" branding in FY2025–FY2026 to capitalize on the LLM/agent wave. [S2]

C3 AI Applications (300+ pre-built applications)

Domain-specific, production-ready AI applications across 8 verticals:

• Energy: Reliability, production optimization, energy management
• Financial Services: Fraud detection, AML, credit risk, CRM
• Government / Defense / Intelligence: Predictive maintenance, logistics optimization, threat detection
• Manufacturing: Predictive maintenance, quality, supply chain
• Healthcare: Patient engagement, chronic disease management
• Telecommunications: Network reliability, churn prediction
• Retail/CPG: Demand forecasting, inventory optimization
• Utilities: Grid reliability, energy efficiency

C3 Generative AI (Gen AI)

Enterprise generative AI applications launched in FY2023–FY2024. Includes enterprise search (retrieval-augmented generation over proprietary data), AI assistants, and multi-agent orchestration. Designed as a bridge between pre-built applications and more open-ended AI use cases. [S2]

4. Revenue Architecture

Revenue Streams

Business model shift: Pre-FY2023, C3.ai derived ~$92M/year (30%+ of revenue) from the Baker Hughes joint venture — a near-captive, high-margin stream. As Baker Hughes wound down (fully by FY2024 at ~$16M), C3.ai had to replace this with organic commercial sales, exposing the true competitive challenges of its direct-sales model. [S3]

Go-to-Market Evolution

• Pre-FY2023: Large enterprise, direct-sales, enterprise license agreements (ELAs); few large deals; heavily dependent on Baker Hughes
• FY2023–FY2025: Transition to Individual Purchase Agreements (IPDs), consumption-based pricing, and partner-led model (Azure co-sell, AWS Marketplace, GCP Marketplace, McKinsey, PwC, Booz Allen Hamilton)
• FY2025–FY2026 (Ehikian era): "Agentic AI" re-pivot; emphasis on federal/national security contracts; Ehikian background in federal AI deployments at AWS Government

FY2025 agreement count: 174 total agreements (vs. 123 in FY2024 and 49 prior) — IPD model driving volume growth but lower ACV. [S2]

Customer Concentration Risk

Historically extreme (Baker Hughes ~30%+ of revenue through FY2022). Post-BH wind-down, remaining top-10 customers likely represent 40–50% of revenue (estimates; exact figures require FY2026 10-K). Federal/DoD customers include USAF, Army, DHS, DoE — increasing government revenue concentration as a replacement for BH dependency.

5. Partner Ecosystem

Partner-driven bookings grew +419% in Q4 FY2024 per investor presentation data, suggesting the pivot is producing pipeline even if it has not yet stabilized revenue. [S4]

6. Competitive Position Summary

C3.ai competes against:

1. Hyperscalers (Azure AI, Google Vertex AI, AWS SageMaker): Far more resources, bundled pricing advantage, but C3.ai claims pre-built apps reduce enterprise deployment time from years to months
2. Palantir: Most direct comparable — both sell enterprise AI to governments and large corporations. Palantir has $3–4B revenue vs. C3.ai's ~$250M, and is now profitable.
3. Specialized AI vendors: Salesforce Einstein, IBM watsonx, DataRobot — narrow focus, less platform breadth

C3.ai's stated differentiation: Production-ready apps, multi-cloud neutrality (runs on Azure/AWS/GCP equally), and the ability to integrate with existing enterprise data infrastructure without a "rip and replace." [S5]

7. Management & Ownership

• Stephen Ehikian (CEO, since September 2025): Former AWS Government head; enterprise AI deployment background; brought in to execute the federal/DoD pivot
• Tom Siebel (Executive Chairman/Founder): 38 Form 4 transactions since 2021 — all sales; systematic liquidation via 10b5-1 plan; retains ~9.1M shares but reduced control
• Dual-class structure: Class B shares (50 votes/share) gave Siebel near-total voting control; post-transition, governance dynamics unclear without FY2026 proxy detail

8. Source Index

source: coverage-next-full ticker: AI company: C3.ai, Inc. step: 04 title: Financial Quality & Adversarial Sweep created: 2026-06-11

Step 04 — Financial Quality & Adversarial Sweep: C3.ai, Inc. (AI)

1. Executive Summary

C3.ai's reported financials contain multiple flags warranting careful scrutiny: SBC exceeding revenue in FY2026, anomalous gross margin collapse, persistent GAAP losses despite a $700M+ cash cushion, and a history of customer concentration that inflated apparent growth metrics. The Adversarial Research Sweep below examines the most serious external allegations and accounting concerns. The core assessment: C3.ai's financial reporting appears technically compliant but paints an overly optimistic picture through non-GAAP exclusions of SBC. The real cash burn is obscured by interest income on the cash hoard. [S1]

2. Statement Quality Adjustments

Income Statement Adjustments

Key insight: GAAP operating loss of ~$152M looks severe, but FCF burn of only ~$47M (FY2025) shows the difference between GAAP (which expensing SBC in full) and cash reality. However, SBC is economically real — it dilutes shareholders. On a cash-adjusted, SBC-included (diluted) basis, the company burns ~$150M+/year economically.

Non-GAAP vs. GAAP: C3.ai reports non-GAAP operating metrics that exclude SBC. Management highlighted "non-GAAP operating loss" improving materially (from ~($90M) non-GAAP to closer to breakeven). This is technically accurate but excludes the company's single largest expense category. [S1]

Revenue Quality Assessment

3. Balance Sheet Quality

Cash Position

Cash declined ~$70M over FY2025 (burn ≈ FCF burn of $47M + other) and another ~$69M over FY2026 YTD (Q3 FY2026 implied). The cash runway is substantial (>5 years at current burn rate), but the direction is negative.

Receivables / DSO

DSO estimation from XBRL not fully available, but professional services contracts and federal contracts can have extended payment terms (60–90+ days). Worth monitoring for FY2026 10-K.

Intangibles / Goodwill

C3.ai has minimal goodwill from acquisitions — the company has grown organically. No major goodwill impairment risk.

Operating Leases

Long-term operating leases (~$34M remaining per FY2025 10-K) for office space. No financial debt. Clean balance sheet from a leverage perspective.

4. Adversarial Research Sweep

Methodology: Web search for short seller reports, SEC investigations, class action lawsuits, whistleblower allegations, regulatory enforcement actions, and major accounting controversies related to C3.ai. No earnings transcripts loaded — relies on public filings, press coverage, and regulatory databases.

Finding 1: Short Seller Reports (Multiple, 2021–2023)

Source: Multiple short-selling research publications; widely covered in financial press [S3] Allegation: In 2021–2022, short sellers (including alleged Spruce Point Capital research coverage) questioned C3.ai's revenue recognition practices, particularly around the Baker Hughes alliance — alleging that the revenue was essentially a "barter" arrangement where Siebel and Baker Hughes CEO Lorenzo Simonelli had interlocking incentives. The allegation was that reported BH revenue was artificially inflated and would collapse once the relationship normalized. Current status: The collapse of BH revenue (from $92M to $16M by FY2024) largely validated the concern — it was a related-party distortion, not organic customer demand. C3.ai disclosed this as a related-party transaction and wound it down properly. No SEC enforcement action resulted. Severity: RESOLVED (the distortion is now in the past). The bear case was right — BH revenue was not reproducible.

Finding 2: IPO Lock-Up / Insider Selling Pattern

Source: Form 4 filings via SEC EDGAR [S4] Allegation: Not an allegation per se, but 38 Form 4 filings for Tom Siebel with zero purchases since 2021 — all sales. Systematic liquidation via 10b5-1 plan. Remaining ownership ~9.1M shares down from original post-IPO holdings. Interpretation: While 10b5-1 plans can be set up in good faith, the consistent selling pattern with no purchases is a negative signal for long-term conviction from the founder. The CEO transition to Ehikian with Siebel remaining as Executive Chairman may be a succession pathway or may indicate Siebel's reduced commitment. Severity: MEDIUM — ongoing overhang; not fraud, but negative governance signal.

Finding 3: Revenue Recognition and IPD Model Integrity

Source: FY2024 and FY2025 10-K risk factor disclosures [S1] Concern: The shift to Individual Purchase Agreements (IPDs) — consumption-based contracts — creates revenue recognition timing variability. Unlike traditional SaaS ARR, consumption revenue can be lumpy. The dramatic Q3 FY2026 revenue step-down (-44% YoY, -29% sequentially) could indicate:

1. CEO transition disrupting sales organization (legitimate operational issue)
2. IPD contracts not renewing / customers consuming less than contracted
3. A federal contract timing gap (contracts delayed by CR/budget uncertainty)
4. Channel conflict between partner-led and direct sales during transition Evidence for/against fraud: No SEC investigation; 10-K disclosures are detailed and disclose all material customer concentration. Audit firm (KPMG) has not flagged material weaknesses. Severity: MEDIUM-HIGH — not fraud, but the Q3 FY2026 step-down needs management explanation (available in earnings call transcript, which is not loaded here).

Finding 4: Class Action Lawsuits

Source: Legal databases; press coverage [S3] Status: C3.ai faced securities class action litigation in 2023 related to alleged misrepresentations about growth prospects and the Baker Hughes relationship. Settlement status as of research date unknown. Management disclosed the litigation in FY2024/FY2025 10-K risk factors. Financial impact: Any settlement would be a one-time charge; no indication of material financial impact. Severity: LOW-MEDIUM (ongoing legal risk, not existential).

Finding 5: SBC > Revenue (FY2026)

Source: StockAnalysis.com FY2026 data; SEC XBRL SBC trend [S2] Concern: If FY2026 SBC is ~$264M on ~$250M revenue, this is economically unsustainable. Large new CEO grants to Ehikian + leadership refresh likely drove this spike. However:

• Grants to new executives are typically front-loaded and amortized over vesting schedules
• Future SBC should decline as grants vest (if no new major refresh occurs)
• The company will need to disclose SBC trajectory in the FY2026 10-K Severity: HIGH (dilution risk) — but may be transient if driven by CEO onboarding grants.

5. Audit Quality Assessment

• Auditor: KPMG LLP (Big 4)
• Audit opinions: Unqualified in FY2025 and FY2024 10-K
• Material weaknesses: None disclosed in recent filings
• Related-party disclosures: Adequately disclosed (Baker Hughes, board members)
• Revenue recognition policy: ASC 606 compliant; subscription and PS revenue recognized separately

Overall audit quality: Satisfactory. No red flags in audit opinions. The Baker Hughes related-party transaction was properly disclosed throughout.

6. Cash Burn Analysis

FCF improving trend (FY2022→FY2025): From -$114M to -$47M — genuine improvement driven by operating leverage. However, FY2026 may reverse this as revenue declines faster than costs.

Cash runway: At $673M cash and ~$65M/year burn rate, C3.ai has ~10+ years of runway — not an existential concern.

7. Source Index