Priced for Perfection

Three records fell in Q1 2026 that don't normally coexist. Global VC hit $300 billion — the highest quarterly total in history — with 80% going to AI. Goldman Sachs measured 16,000 net US jobs lost to AI per month, with AI cited in 25% of all March layoff notices, up from 10% in February. And 90% of enterprises still report no measurable productivity impact from GenAI, per NBER. Capital is being committed at the fastest pace in history based on returns that 9 in 10 enterprise users can't yet measure. The bet is that they eventually will.

1 predictions updated2 milestones0 companies refreshed

Key Developments

OpenAI raised $122 billion in a single quarter at an $852 billion valuation — more than ExxonMobil, JPMorgan, and Visa combined

OpenAI's Q1 2026 round is the largest private fundraise in history, with Amazon and Nvidia among backers, pushing its valuation to $852 billion — approaching Apple-scale for a company still spending more than it earns. Anthropic followed at $380 billion (with revenue reportedly now surpassing OpenAI's) and an IPO targeting October 2026. But strip the five largest deals from Q1's record $300 billion in global VC, per PitchBook, and the rest of the market raised just $72 billion — with seed deal count down 30%. The headline says boom. The data underneath says concentration.

Challenges P-012 ↘ AI investment boom is a bubble likely to burst by ...

Counterpoint

The bull case isn't hypothetical. Amazon's $200 billion 2026 CapEx commitment (+60% YoY) flows from record AWS margins, not future promises. Meta spent $65 billion and grew operating income 50%. Concentration at the top — twelve firms capturing 73% of AI VC — isn't necessarily a bubble; it may be an oligopoly forming around structural network effects. OpenAI losing money at $852B is different from WeWork losing money at $47B: the revenue trajectory is real, even if the valuation multiple requires everything to go right.

Source →

A federal judge called the Pentagon's Anthropic blacklist 'Orwellian' — and issued an injunction

Federal judge Rita Lin granted a preliminary injunction blocking DoD from enforcing its "supply chain risk" designation against Anthropic — the designation Secretary Hegseth imposed after Anthropic refused to enable AI-driven domestic surveillance and autonomous weapons. The ruling: no statute authorizes labeling an American company a national security threat for disagreeing with a procurement officer's policy demands. Three weeks earlier, Microsoft, Google, and OpenAI all filed court documents opposing the designation — competitors rallying behind a peer because they couldn't afford the precedent to stand.

Counterpoint

A preliminary injunction establishes that Anthropic has a plausible legal argument worth a full hearing — not that the argument will succeed at trial. Courts give the executive branch wide deference on national security procurement decisions, and DoD lawyers will argue exactly that. More practically: the Pentagon doesn't need a formal blacklist designation to stop buying Claude. It can simply not renew contracts. The injunction prevents the label; it doesn't require a signature.

Source →

8 in 10 Fortune 500 companies now run AI agents — and Agentforce just hit $800M ARR, up 169% in a year

State of AI Agents 2026 puts active agent deployment at 80% of Fortune 500, with 57% running multi-step autonomous workflows — ahead of Gartner's 40% end-2026 threshold by any practical measure. Salesforce validated the revenue model: Agentforce hit $800M ARR and relaunched Slack as a natural-language interface to all Salesforce products, powered by Anthropic, aimed directly at Microsoft Copilot's installed base. NVIDIA added an Enterprise AI Agent Toolkit at GTC with 17 named adopters including Adobe, Salesforce, and SAP — creating common infrastructure for agents across cloud and on-premise deployments. The seat compression thesis is now generating more revenue than it's destroying, at least at Salesforce.

Confirms P-021 ↗ 40% of enterprise applications will feature task-s...

Counterpoint

The 80% Fortune 500 adoption figure and the 14% "actually at production scale" number (Deloitte, March 2026) can both be true — and are. Writer's 2026 global survey of 2,400 executives found 79% of enterprises face significant AI adoption challenges; 54% of C-suite leaders say it's tearing their organization apart. Agentforce's $800M ARR is real. So is the 95% GenAI enterprise pilot failure rate. The gap between "active agent adoption" and "agents doing work that matters at scale" is the central unresolved question of 2026.

Source →

What the Evidence Moved

P-020AI-exposed industries will see job losses of ~20,000 per mon...

Goldman Sachs publishes first rigorous AI displacement decomposition: 16K net US jobs/month (25K gross substitution, 9K new augmentation roles). Q1: 80K tech layoffs with 50% AI-cited — first quarter crossing 40% threshold. Oracle April wave (20-30K) will push April pace above monthly record. Direction confirmed; rate now within range of Goldman's original 20K/month projection.

55%→60%▲ +5pp

Company Impact

Salesforce

Data refresh

Agentforce $800M ARR (+169% YoY). Claude-powered Slack relaunch targeting Microsoft Copilot. NVIDIA GTC Agent Toolkit adopter. Score update recommended in next batch.

Fortune · VentureBeat

6.7

See assessment →

OpenAI

Data refresh

$122B raise at $852B valuation — largest private fundraise in history. IPO targeting Q4 2026. Revenue reportedly below Anthropic now.

Crunchbase · Techi

9.1

See assessment →

Anthropic

Data refresh

$380B valuation, revenue reportedly surpasses OpenAI. IPO possible October 2026. Pentagon blacklist blocked by federal judge (preliminary injunction, DoD can appeal).

TradingKey · CNN

9.1

See assessment →