// Monday · November 16, 2026

5 Questions That Could Reshape AI

Anthropic prices the most consequential AI IPO yet while the Jensen Huang saga reverberates from Wall Street to the Situation Room. Five questions will shape this week: how the IPO goes, whether OpenAI or a Chinese lab tries to steal the spotlight, whether Jensen comes back, whether Howard Lutnick takes the fall, and whether Congress finally moves on export controls.

Ad-free on Patreon
Sponsored by KPMG · Robots and Pencils · CData · Blitzy · all offers →
The One Idea

This is a before-and-after week for AI, and Anthropic's IPO is the hinge.

Anthropic's IPO was always going to be a landmark. Add last week's Jensen Huang scandal and it becomes a moment the history books will mark. For years, investors have used Nvidia, the neoclouds and stock-market machinations as bad proxies for the only question that matters: will AI demand justify the cost? Once Anthropic is public, every quarter brings fresh evidence for that debate. Around it swirl four more questions that will shape far more than AI: OpenAI's next move, Jensen's possible return, Lutnick's exposure and the fate of export controls.

// 01

By the Numbers

$1.25–1.4T
Anthropic's valuation target as of last Monday, with some bankers pushing for $1.5T
$90B+
Total raise including IPO, private placement and greenshoe
~7×
Oversubscription heading into the weekend, reportedly hotter than SpaceX
$900B+
Anthropic's disclosed compute commitments over the next decade
$71B
Compute commitments coming due over the next year
72 days
Since OpenAI's last frontier model release
175,000
Signatures on the open letter to bring back Jensen, including 10,000+ employees
51%
Gene Munster's odds of Jensen's return, up from 10% (Kalshi: 9%)
1.5GW
Nvidia chips in China that shouldn't be there, per SemiAnalysis
// 02

The Brief

BusinessFinanceExec▶ 02:00

Anthropic is about to become the world's most important AI stock

Bloomberg's headline captures the stakes. Expectations for markets, the American economy and even geopolitics are bundled into this one IPO. First trading is slated for Thursday the 19th, with final pricing the day before.

AI Daily Brief
BusinessFinance▶ 02:00

From $2 trillion dreams to a $1.4 trillion target, and bankers pushing back up

Peak-bullish investors once talked about a $2T valuation. As the AI trade soured, the target fell to $1.25–1.4T, which some banking partners reportedly plan to push to $1.5T. The raise is $57–64B from IPO buyers plus $26B in private placement led by Nvidia and Google, more than $90B all in. Anthropic reportedly cares more about cash raised than about the perfect valuation.

AI Daily Brief
BusinessFinance▶ 03:00

Demand is running nearly seven times supply

The deal was reportedly six times oversubscribed last Thursday and closing on seven by the weekend. One source called demand hotter than SpaceX at this stage and compared it to Snowflake's 2020 IPO, which popped 112% on day one. The difference is scale: Anthropic is raising three times Snowflake's entire valuation at the time.

AI Daily Brief
BusinessFinance▶ 04:00

The cynical case: Anthropic can't afford a broken IPO

Scott Galloway argues that whatever the fundamentals, the first trade is almost certain to rip. He expects it 15–40% above pricing because Anthropic would rather take more dilution than face a 'broken IPO' headline. His risk-adjusted play is to get an allocation and sell on the first trade.

AI Daily Brief
ComputeFinanceExec▶ 05:00
“

Compute is the deepest business dilemma that the company faces.

— Dario Amodei, Anthropic CEO, at the company's investor day. At a JPMorgan investor day hosted by Jamie Dimon, Amodei said the field is moving so fast technologically that nobody knows how it will translate into economic value, which leaves Anthropic with 'huge uncertainty.' Wall Street is struggling to value a company that fits no standard tech profile.

The AI Daily Brief
ComputeFinanceExec▶ 06:00

Anthropic is deliberately overbuying compute, and $71B comes due this year

Anthropic has disclosed more than $900B in compute commitments over the decade, with $71B due in the next year. Amodei said an earlier bet on too little compute cost the company customers and R&D cycles and stunted revenue growth. Buying too much carries its own risk: financial overextension.

AI Daily Brief
◆ The TakeFinanceExec▶ 08:00

Nvidia was always a terrible proxy for the real question

Analysts call Anthropic the new bellwether, maybe more important than Nvidia now that Jensen is gone. NLW goes further. Even with Jensen still there, everything investors watched failed to answer the only question that matters: will AI demand justify the cost? Once Anthropic is public, every quarter brings real evidence for that debate.

The AI Daily Brief
ModelsProductEng▶ 08:00

The overlooked story: Anthropic is shipping whole apps at a blistering pace

Last week alone brought a productivity suite and then a video-meetings app. Just as Opus 4.5 let Claude Code ship features almost daily, internal models may now be letting Anthropic ship entire products. It is chipping away at nearly every major software category.

AI Daily Brief
PolicyLegalExec▶ 09:00

David Sacks accuses Anthropic of building the world's most valuable monopoly

The former AI czar says Anthropic's 'winner-take-all' roadshow language describes a monopoly. He also charges that the company calls recursive self-improvement the most dangerous work there is while marketing it to investors as a business advantage.

AI Daily Brief
BusinessFinance▶ 10:00

The roadshow pitch: end disease, then everything else

Amodei told investors he hopes heart disease, Alzheimer's and cancer become relics like smallpox: 'It sounds crazy. I really do believe it.' One account of a pre-IPO broker call went further. It described a pitch about taking on hyperscalers, every other lab, biotech and robotics, with a $1T ARR target by Q4 2028.

AI Daily Brief
ModelsExec▶ 10:00

Will OpenAI try to steal Anthropic's IPO spotlight?

With the IPO now looking safe, and Amodei taking potshots ('We don't vague post about models ahead of time'), OpenAI has reason to ship. Leaker Token Gremlin showed outputs said to come from a GPT-6.2 Astra preview. They also reported that API resellers already have the model, which, if true, means a release is close.

AI Daily Brief
ModelsEng▶ 11:00

Chinese models don't need to be good to spook Wall Street

A Chinese release only has to look strong long enough to drive a narrative cycle. Moonshot and DeepSeek look unlikely right now. Qwen 4 Max claims benchmark wins over Opus 5.5 and GPT-6.1, but real-world tests look lackluster. One developer's diagnosis: since Anthropic shut the door on distillation, Chinese models 'haven't moved an inch.'

AI Daily Brief
Models▶ 15:00

OpenAI's frontier-model drought hits 72 days

That is the longest gap since GPT-5 to 5.1. A few more weeks without a release would make it the longest since o3 to GPT-5. Whether or not a model lands this week, something has to come soon.

AI Daily Brief
◆ The TakeExec▶ 16:00

Is OpenAI in third place? Not quite consensus, but bubbling

A growing narrative says OpenAI has fallen behind. NLW agrees that Opus 5.5 is ahead. He thinks Google, despite a strong Gemini 4 Carbon, still has more to prove before it can claim second.

The AI Daily Brief
PolicyLegal▶ 16:00

A possible reason for the delay: Washington now gates model launches

OpenAI's Rune says people underestimate federal involvement: a new model must be shown to be safe to the government before launch, or export-control bans follow. NLW's bet is a new OpenAI model soon after the Anthropic IPO, with Thanksgiving complicating the timing.

AI Daily Brief
BusinessExec▶ 17:00

The Bring Back Jensen movement crosses 175,000 signatures

A new open letter counts more than 10,000 current and former Nvidia employees and 90,000 shareholders among its signers. The effort reportedly has Jensen's support. Shareholders Gavin Baker and Brad Gerstner also wrote to the board urging a deliberate process that keeps his return on the table.

AI Daily Brief
BusinessExecLegal▶ 18:00

'Resigned,' not fired, gives the board a way back

The New York Times reports that Trump urged Jensen to fight for his job, seeing the economy as tied to his personal leadership. Gene Munster raised his odds of a return to 51% from 10%, while Kalshi sits at 9%. Because Jensen technically resigned, the board can welcome him back without admitting it fired him. Balaji Srinivasan argues it is a fait accompli if Jensen simply asks.

AI Daily Brief
◆ The Take▶ 19:00

The mainstream sees a sex scandal; the real story is the chips

SNL framed the saga as a midlife crisis that tanked markets, plus a skit about the CIA teaching billionaires to spot spies. NLW points out what all of it misses: the security implications of Jensen potentially knowing that chips were reaching China.

The AI Daily Brief
Policy▶ 20:00

The spy angle is getting murkier

DealBook accused David Sacks of spreading conspiracy theories, but he was amplifying CNN reporting that CIA and DOJ officials aren't certain Alice Huang was a Chinese spy. Taylor Lorenz defended him. Adding another curveball, Megaspeed says she left the firm in August of last year.

AI Daily Brief
PolicyLegalExec▶ 22:00
“

It doesn't matter to me whether she's a spy or not.

— Treasury Secretary Scott Bessent, on the Sunday shows. Bessent wants the board to look outside the company for a 'dramatic shakeup.' He calls the real question whether Nvidia's culture pushed seemingly illicit sales to pass-through shell companies to goose revenue. He wants answers from the board and outside counsel within 10 days.

The AI Daily Brief
ComputeLegal▶ 22:00

SemiAnalysis: 1.5GW of Nvidia chips are in China illegally

Dylan Patel argues Nvidia definitely knew. He says unwinding the web of how those chips got there is what makes this dangerous, and why the board moved so fast. Whether the board acted to stave off investor lawsuits or because intelligence agencies came asking determines whether Jensen's return is even in its hands.

AI Daily Brief
BusinessFinance▶ 23:00

Tuesday's earnings won't be the story; the China questions will

Nvidia has beaten estimates 15 quarters in a row and likely will again. Interim CEO Colette Kress should expect early questions about chips going to China. Dan Ives sees 'no chance' Jensen returns, arguing the rockstar persona is gone for good.

AI Daily Brief
PolicyLegal▶ 24:00

Lutnick emerges as the scandal's central Washington figure

Lutnick reportedly stalled three times on cracking down on Megaspeed. The NYT says Nvidia's general counsel inserted comments directly into Commerce policy drafts. He testifies before Senate Banking on December 10. Bessent says only that 'Commerce has the pen,' and some Democrats say the scandal goes all the way to the president.

AI Daily Brief
PolicyLegalExec▶ 26:00

With Jensen gone, who's left to lobby against export controls?

Three export-control bills, including one closing the remote-access loophole for Southeast Asian data centers, are headed into must-pass NDAA negotiations. Speaker Mike Johnson sees a 'developing consensus.' With both OpenAI and Anthropic backing tougher controls, the status quo has almost no defenders left.

AI Daily Brief
Machine-readable ▸Download .mdTranscript .md— feed it to your own agent

Got this from a colleague? Get the brief every day.