Edward Zitron is an independent technology journalist and media analyst, and author of the newsletter Where's Your Ed At (wheresyoured.at). He is among the more persistent critics of the AI industry's financial claims and capability narratives.
Positions and statements
On the economics of AI labs, Zitron argues that the major labs are losing billions of dollars and subsidizing compute for users at a rate of $3 to $15 per dollar of subscription revenue, and that the underlying economics are structurally broken. He has characterized the sector as an "AI bubble."
On capabilities, Zitron argues that every product marketed as an "AI agent" reduces to "a chatbot connected to an API," that current agents fail at complex tasks, and that media coverage has created a false narrative of autonomous AI.
On revenue reporting, Zitron has reported on what he describes as discrepancies in OpenAI and Anthropic revenue figures, including the contrast between a CFO affidavit and annualized revenue claims. He disputes the framing of AI training costs as "R&D" or capital expenditure, arguing instead that training is an ongoing cost of goods sold (COGS).
Zitron's strongest claims, such as the argument that Anthropic and OpenAI may be misleading investors, remain unverified absent S-1 filings.
Notable work
Zitron's work sits at the intersection of technology journalism and financial analysis, and he reports findings that larger outlets often ignore or contradict. His writing on AI includes:
- "AI Is Really Weird" (April 8, 2026), a skeptical analysis (AI Is Really Weird).
- "The Subprime Data Center Crisis" (July 22, 2026), arguing that AI data-center financing is accumulating subprime-style systemic risk, comparing layered debt claims on the same underlying assets to the synthetic CDOs of the 2008 crisis (Source: wheresyoured.at). See Private Credit & AI Infrastructure.
- Hater's Guide to the AI Bubble, a premium newsletter series.
- Hater's Guide to Anthropic (February 2026, premium).
- An investigation of OpenAI revenue discrepancies against Microsoft revenue-share documents (November 2025).
His November 2025 OpenAI revenue investigation was picked up by only a few outlets and then largely set aside by media that continued to report a $13B 2025 revenue figure. His analysis of Anthropic's accounting methods, including non-GAAP margins and annualization methodology, draws on The Information, Newcomer, and official Anthropic announcements.
Relationships
- related: Anthropic — critical coverage of revenue claims and business model
- related: OpenAI — financial investigation, revenue reporting discrepancies
- instance-of: AI as Normal Technology — shares skepticism about AI capabilities and timelines, though from a financial and business angle rather than an academic one