Altman: Musk Demanded OpenAI Control.

Mazen Al Sabbagh

— 7 min read
Share

TL;DR

  • Altman Testimony: Sam Altman claimed Elon Musk demanded "total control" of OpenAI, including generational ownership transfer. Musk's counsel countered, leveraging former OpenAI executives' testimony and Altman's financial conflicts to challenge his credibility. Closing arguments are scheduled for Thursday.
  • Anthropic Valuation: Anthropic is reportedly seeking $30–50 billion in funding, potentially at a $900–950 billion valuation. This maneuver includes a new legal-tech vertical and acquisition plays, strategically positioning the company ahead of a potential IPO.
  • Seoul Market Jolt: A South Korean policy chief's "AI profit dividend" suggestion triggered a $300 billion flash crash on the Kospi. Though clarified as personal opinion, the incident underscored market sensitivity to AI wealth redistribution proposals.
  • Huang's Beijing Pivot: Nvidia CEO Jensen Huang joined President Trump’s delegation to Beijing. This unplanned inclusion places AI chip export policy directly onto the summit agenda, impacting global tech and trade relations.
  • Cerebras IPO: Cerebras is guiding its IPO pricing above its initial range, with 20x oversubscription, targeting a ~$34.4 billion valuation. This positions it as 2026’s largest IPO, signaling robust institutional appetite for AI hardware.

Lead Story: Altman vs. Musk — The Final Testimony

Sam Altman's four-hour testimony Tuesday provided the most pointed exchanges in the ongoing Musk v. OpenAI trial. Called by his own defense, Altman asserted that Musk was "obsessed" with controlling OpenAI, demanding both profit participation and the right to convey his ownership stake to heirs. This framing challenges Musk's narrative of having been misled into donating to a non-profit that subsequently transitioned to a for-profit entity, a move Altman described as fundamentally difficult to reconcile with Musk's public statements.

Cross-examination by Musk's lead attorney quickly shifted focus to Altman’s personal credibility. After questioning his trustworthiness, the defense methodically presented testimony from former OpenAI insiders, including Mira Murati, Ilya Sutskever, and Helen Toner, who had previously indicated patterns of dishonesty or manipulation. Further scrutiny arose from Altman's financial interests in companies such as Stripe, Cerebras, and Helion, which conduct business with OpenAI. These financial stakes, particularly in Helion, have also drawn congressional attention, with a House Oversight chair requesting disclosures.

Altman's defense included the claim that Musk attempted to undermine OpenAI on two occasions: during the 2023 board crisis and again through the current lawsuit. He also acknowledged considering a move to Microsoft during his temporary ouster, a decision driven by significant financial incentives. Despite the courtroom drama, the trial has yet to yield a substantive dialogue on AI governance, suggesting its scope remains largely a dispute over control rather than a policy reckoning.

Witness testimony concludes Wednesday, with closing arguments scheduled for Thursday. An advisory jury verdict is anticipated as early as next week. Musk's petition seeks up to $150 billion in disgorgement, the removal of Altman and Brockman from leadership, and the dissolution of OpenAI's for-profit structure. OpenAI's current court-filed valuation stands at $852 billion.

In Other News

Anthropic Targets $950B Valuation in Mega-Round. Late Tuesday, reports from the NYT's Mike Isaac indicated Anthropic is discussing a new funding round of $30–50 billion, potentially valuing the company at $900–950 billion and positioning it to exceed OpenAI's valuation. Bloomberg corroborated the $30 billion minimum. This follows a $30 billion Series G at $380 billion months prior. Anthropic has also moved to invalidate unauthorized secondary share trades, tightening control over its cap table in advance of an anticipated October 2026 IPO. Such a round would establish Anthropic as the most valuable private company on record. Concurrently, Anthropic launched "Claude for Legal" on Tuesday, integrating 12 practice-area plugins and over 20 MCP connectors with platforms like DocuSign and Westlaw. This legal vertical's introduction reportedly contributed to a selloff in software stocks.

_Why it matters:_ Anthropic’s aggressive capital raise, strategic acquisitions of developer infrastructure, and rapid vertical expansion are clear indicators of a calculated IPO strategy. This velocity aims to redefine its market position and valuation hierarchy within the AI landscape.

Seoul's AI Tax Trial Balloon Triggers $300B Flash Crash. South Korea's presidential policy chief, Kim Yong-beom, proposed on Facebook Tuesday a "national dividend" for citizens, financed by AI profits. The Kospi index responded with a 5.1% drop, erasing over $300 billion in market capitalization, with Samsung Electronics falling nearly 6%. Kim quickly clarified his remarks referred to reallocating "excess tax revenue" from the AI boom, not instituting a new windfall tax. A presidential office official further distanced the government, stating the remarks were personal. The Kospi subsequently recovered, closing up 2.63% at a new record of 7,844 as investors engaged in bargain hunting. Sympathetic selloffs and recoveries were also observed in international memory and chip stocks.

_Why it matters:_ The incident starkly illustrates market sensitivity to government intervention, however informal, in the distribution of AI-generated wealth. The rapid market reaction underscores the financial leverage held by the concentrated profits within the AI sector.

Jensen Huang Boards Air Force One for Beijing. In an unexpected development, Nvidia CEO Jensen Huang received a direct invitation from President Trump on Tuesday evening to join the presidential delegation to Beijing. Huang subsequently boarded Air Force One in Alaska during a refueling stop. This insertion of Huang into the summit itinerary elevates AI chip export policy to a primary agenda item alongside established discussions on trade, Iran, and Taiwan. The delegation now includes Huang, Tim Cook, Elon Musk, and Larry Fink. Politico reports the summit's broader ambitions have been scaled back to a narrower agenda, largely influenced by the Strait of Hormuz crisis and China's "four red lines" concerning Taiwan, sanctions, and tech restrictions.

_Why it matters:_ Huang's presence fundamentally shifts the summit's dynamic, transforming it into a direct negotiation on technology trade and supply chain access. Any adjustment to AI export controls resulting from these talks would immediately restructure the global AI market.

Cerebras IPO Set to Be 2026's Biggest. Cerebras is reportedly guiding its IPO pricing above its already-raised $150–160 range, with the offering currently 20x oversubscribed. This trajectory positions the company for a ~$34.4 billion valuation at $160 per share, aiming to raise up to $4.8 billion. Such figures would establish it as 2026's largest IPO. Wall Street's enthusiasm has been characterized as a search for "the next Nvidia." Cerebras is part of a series of AI-adjacent IPOs this week, alongside Fervo Energy and a Blackstone data-center REIT, collectively raising approximately $8 billion. Notably, Sam Altman holds a personal stake in Cerebras, a detail highlighted during Tuesday's trial testimony concerning potential conflicts of interest.

_Why it matters:_ The market's aggressive embrace of the Cerebras IPO confirms a sustained and substantial institutional demand for direct exposure to AI hardware, consolidating the nascent public market for specialized AI chip companies beyond Nvidia.

X / Social Pulse

On Altman's testimony: The cross-examination's opening question, "Are you completely trustworthy?", dominated the AI-focused social discourse. Legal commentators largely observed that Musk's counsel effectively leveraged testimony from OpenAI's own former executives to question Altman's character. Conversely, Altman's assertion that Musk "tried to kill OpenAI twice" gained traction as a potentially memorable soundbite for the jury.

On the Korean flash crash: Macro and AI traders expressed surprise at the speed and scale of the Kospi's selloff. Many noted this marks a rare instance where an AI-specific tax proposal, even an informal one, immediately triggered a multi-hundred-billion-dollar movement in global equity markets. The subsequent recovery was attributed to the official clarification and rapid bargain hunting, but the underlying market sensitivity is now a priced-in factor.

On Anthropic's valuation: The reported $950 billion valuation for Anthropic elicited significant discussion. A segment of the VC community questioned the justification for a near-trillion-dollar valuation for a company not yet generating revenue at scale. Others pointed to Anthropic’s substantial compute commitments across Google Cloud, Akamai, and SpaceX as evidence of proportional spending that anticipates significant future revenue streams.

One to Watch

Trump-Xi Summit Day One — AI at the Table. President Trump is now in Beijing, with Jensen Huang present at the negotiating table. China has outlined its "red lines." The two-day summit begins Thursday, formally including AI on the agenda alongside Iran, trade, and Taiwan. Reports suggest a focus might lean towards AI-enabled warfare, particularly following its deployment in Gaza, rather than solely on semiconductor export controls. Huang's participation, however, renders chip policy unavoidable. Stanford's latest AI Index indicates the performance gap between US and Chinese AI models has "effectively closed." The outcome of this meeting will define the trajectory for AI governance between the world's two dominant economies. Domestically, White House efforts to finalize its own AI executive action are reportedly stalled by internal disagreements, leaving the US with a fragmented policy stance entering this high-stakes summit.

Quick Hits

  • Isomorphic Labs Funding: DeepMind spinout Isomorphic Labs, led by Demis Hassabis, closed a $2.1 billion Series B round. Participants included Thrive Capital, Alphabet, Abu Dhabi's MGX, Singapore's Temasek, and the UK Sovereign AI Fund.
  • Meta's WhatsApp Offer: Meta extended a one-month free access offer for WhatsApp to rival AI chatbots, including Claude and ChatGPT, a preemptive measure to mitigate potential EU antitrust penalties.
  • Samsung Market Rout: Samsung Electronics experienced an intraday loss of up to $66 billion Wednesday as the Suwon District Court concluded hearings on Samsung's injunction to block a May 21 strike; a ruling is pending.
  • xAI Sues Colorado: xAI filed a lawsuit to block the enforcement of Colorado's AI Act. The Department of Justice intervened, resulting in a suspension of the act’s enforcement.
  • Japan's Mythos Access: Japan's three largest megabanks are expected to gain access to Anthropic's controversial, cyber-capable Mythos model within two weeks, marking its first significant deployment within the financial sector.

The trial involving Altman and Musk enters its concluding phase Thursday with closing arguments. Concurrently, the valuation narratives continue their ascent: Anthropic at $950 billion, OpenAI at $852 billion, and Cerebras poised to price its IPO above $34 billion. Geopolitically, Jensen Huang’s presence with Trump in Beijing signals a direct integration of AI hardware into top-tier trade discussions. Domestically, Seoul’s accidental market jolt demonstrates the immediate and substantial financial impact of policy discussions surrounding AI wealth distribution. The capital is immense, the strategic stakes are global, and the transparency regarding these realities is now unequivocal.

Sources

Lock in. M. mazen@thorterminal.com

Read more