Price:

POLITICS

The White House seeks equity stakes in OpenAI and Anthropic

Monday, June 29, 2026 · from 4 podcasts, 7 episodes
The a16z ShowAI Daily BriefThe DailyBreaking Points
  • The U.S. government pivoted from voluntary AI safety reviews to negotiating ownership stakes.
  • Startup founders warn nationalization would create CCP-style corporate-government fusion.
  • The shift is opening a door for China’s open-source AI to undermine American dominance.

The era of permissionless AI release ended on June 26. Sam Altman confirmed GPT-5.6 would not see a general public release because the government requested a 'limited preview' instead. Under this new arrangement, the White House is effectively approving access on a customer-by-customer basis.

Nathaniel Whittemore described the shift as an informal, technically incompetent licensing regime. Zvi Mowshowitz called it 'maximally terrible' because it relies on arbitrary, non-transparent requirements rather than established law. The friction isn't just about safety; it's about control.

The next day, President Donald Trump confirmed reports that the government is exploring an equity stake in major AI labs. He suggested a plan where the American public becomes partners in these companies. Sam Altman had already met with Bernie Sanders to pitch donating equity to seed a national wealth fund. Trump views this as a PR win, suggesting direct AI dividends for citizens.

"Donald Trump confirmed reports that the government is exploring an equity stake in major AI labs, proposing a concept where the American public could become partners with these companies and benefit from AI success."

- Nathaniel Whittemore, The AI Daily Brief

The logic bridges the aisle. Sanders wants to tax the 'job apocalypse.' Trump wants the public to benefit from the success of the tech giants. Critics like David Sacks warn this fusion creates a CCP-style social credit risk. Brad Gerstner opposes nationalization, citing risks of crony capitalism and a slippery slope.

The move began with a security scare. On June 22, Senator Mark Warner reported that Anthropic’s Mythos model breached almost every classified NSA system within hours. That report, clarified later as a controlled red team exercise, provided the pretext for export controls and the broader regulatory pivot.

Marc Andreessen warns the West’s move toward heavy restriction creates a dangerous paradox. The democratic West is moving toward heavy regulation and restriction, while the Chinese Communist Party is championing open-source AI. He views this as a deliberate 'turbo dumping' strategy. By flooding the world with free, high-quality models, China hopes to destroy the profit margins of American AI companies.

"He notes China's strategic promotion of open-source AI acts as a 'turbo dumping' strategy to flood the market and undermine American commercial viability, creating an ironic dynamic where the 'totalitarian' regime pushes openness."

- Marc Andreessen, The a16z Show

The bottleneck at OpenAI and Anthropic is creating a massive opening for those alternatives. While the feds scrutinize GPT-5.6, Google’s Gemma 4 has already surpassed 200 million downloads. Smaller organizations are flocking to z.ai’s GLM 5.2 model to avoid the uncertainty of the US regulatory environment.

Andrew Curran states that model delays only slow public releases, not training speed, which widens the gap between public and lab-internal AI capabilities. This creates a dangerous pressure cooker. If the public and the 'unfree world' are left behind while the frontier advances in secret, it breaks the Pax Technologica of the last decade.

The debate is no longer about whether to regulate AI, but who will own it.

#Regulation#Big Tech
Nathaniel WhittemoreSam AltmanAnthropicOpenAIChinaGoogleGPT-5National Security AgencyWhite House

Source Intelligence

- Deep dive into what was said in the episodes

The a16z Show
The a16z Show

Beyond P(doom): Marc Andreessen - Betting on America • Jun 29

  • Marc Andreessen argues AI can serve as a 'world's best' tutor, doctor, lawyer, or accountant in your pocket, but current policy prevents it from performing these licensed roles.
  • Research suggests AI boosts productivity for both top performers and median performers, raising the average skill level across fields like law, screenwriting, and programming.
  • Andreessen describes a bifurcated economy: 'blue' sectors (tech, software, TVs) see rapid innovation and price deflation, while 'red' sectors (healthcare, education, housing, law, government) have zero or negative productivity growth and skyrocketing prices.
  • He argues heavy regulation in red sectors restricts supply and subsidizes demand, causing prices to spiral and allowing those sectors to 'eat the entire economy,' suppressing overall growth despite rapid technological change.
  • Alpha School demonstrates a private AI-driven education model where AI handles two hours of academic instruction and teachers focus on six hours of project-based work, but Andreessen believes the public system will resist this change.
  • Andreessen contends 99% of constraints on AI infrastructure are domestic, like county-level opposition to data centers and false memes about water usage, not external tariffs.
  • He frames the U.S.-China AI race as a choice between two contradictory goals: exporting AI for global supremacy or restricting it for safety, with Europe having 'suicidally' regulated itself out of contention.
  • He notes China's strategic promotion of open-source AI acts as a 'turbo dumping' strategy to flood the market and undermine American commercial viability, creating an ironic dynamic where the 'totalitarian' regime pushes openness.
  • Given deep civil-military fusion in China, Andreessen acknowledges the risk of dual-use but argues controls are futile because AI models are just files on a hard drive and U.S. companies lack the counterintelligence to prevent leakage.
  • Andreessen points to a reindustrialization push in defense and energy, with startups in new nuclear, rare earth processing, and U.S.-built transformers, supported by current administration policies and potentially creating a second industrial 'Silicon Valley' around Los Angeles.
Also from this episode: (4)

AI Infrastructure (1)

  • Physical bottlenecks at every layer - energy, data center facilities, turbines, transformers, cooling systems, NVIDIA GPUs, and memory chips - constrain AI development and may halt price declines for intelligence.

Models (2)

  • Advanced AI models like Mythos present a dual-use dilemma: they are superior tools for both cyber attack and defense, creating a policy tension between restriction and rapid deployment for security.
  • Andreessen advocates for maximum export of American AI, aiming for a world where even China runs on it, and using advanced models to armor systems against cyber attacks, including ransomware.

Startups (1)

  • He states successful companies in this space organize around larger national goals like American manufacturing, which attracts talent and co-locates R&D, rather than making a direct financial trade-off against outsourcing.
#AI Infrastructure#Macro#Regulation#China#Open Source
The AI Daily Brief: Artificial Intelligence News and Analysis
The AI Daily Brief: Artificial Intelligence News and Analysis

Nathaniel Whittemore

The Capability Overhang Playbook • Jun 28

  • Nathaniel Whittemore defines the 'capability overhang' as the gap between the latent power of existing models and the real value most individuals and organizations extract from them.
  • Whittemore asserts a forced AI pause is underway due to stalled frontier model releases: GPT-5.6, Claude Sonnet 5, and Gemini 3.5 Pro have been delayed, while Fable 5 remains blocked.
  • Leo from SynthWave reported GPT-5.6's new target release is mid-July and DeepMind delayed Gemini 3.5 Pro due to dissatisfaction with its current state.
  • AI Battle data shows the current wait for GPT-5.6 is 61 days, exceeding previous update gaps of 29, 56, and 49 days within the GPT-5 era.
  • Policy advisor Dean Ball argues the entire US AI industry is frozen from new public releases until the government resolves the Fable situation.
  • Whittemore's Capability Overhang Playbook first advises individuals to create a personal learning agenda by honestly assessing their weaknesses in AI tools and workflows.
  • He recommends building a personal benchmark or eval portfolio: reusable task sets with prompts and success criteria to quickly gauge new model performance.
  • WorkAI Institute Glean study found knowledge workers spend about 2.4 hours weekly organizing context for AI agents, a drain on productivity.
  • To reduce context overhead, Whittemore suggests building portable context assets, either broad-based personal portfolios or per-project context packs.
  • He cites two resources for this: his own project ContextPortfolio.ai and Jim Sanguine's 'The Librarian,' an agentic OS curator.
  • Whittemore advises users to experiment deeply with current AI harnesses by building the same project in both Claude Code/Cowork and Codex to compare interfaces and tool interactions.
  • He recommends exploring specific plugins within tools like Claude Code to discover new capabilities relevant to your role, as experimentation often falls off daily to-do lists.
  • For holdouts, Whittemore urges building a full end-to-end agent architecture, using resources like the free AgentOS program and employing a 'two window' method with a build window and a tutor chat.
  • Whittemore argues individuals should explore model independence using routers like Open Router and open models from Hugging Face, and question their own priorities around cost, privacy, and control.
  • For organizations, he suggests reviewing learning resources and incentive structures for AI adoption, ensuring they reward effective use and sharing of reusable systems.
  • Whittemore warns organizations about an 'overly strong known ROI bias' from token efficiency, which could prioritize efficiency AI over opportunity AI for new products and capabilities.
  • He proposes organizations develop a measurement philosophy linking AI usage to both individual and business outcomes, differentiating between adoption, usage, and outcome metrics.
  • An advanced pattern involves shifting from actively managing AI prompts to architecting loops where AI iterates towards a set goal, utilizing the '/goal' feature as a new primitive.
  • He advises packaging recurring capabilities as reusable 'skills' to make agent work transportable across projects, referencing a past show with Nufar Gaspar on agent skills.
Also from this episode: (2)

Other (1)

  • Prediction market odds for a GPT-5.6 release this week collapsed from nearly 90% to below 30% on Tuesday, indicating a sharp change in expectations.

AI Infrastructure (1)

  • Whittemore recommends turning context portfolios into MCP servers to increase portability and efficiency, gaining familiarity with a key part of the agentic ecosystem.
#Agents#Enterprise#Models#Regulation#AI Infrastructure

The Ad Hoc AI Licensing Regime • Jun 27

  • Nathaniel Whittemore reports that Senator Mark Warner conveyed an NSA finding that Mythos demonstrated significant capabilities during a red teaming exercise, which some initially misinterpreted as the AI breaking into classified systems.
  • Nathaniel Whittemore highlights a new ad hoc, informal, and unaccountable licensing regime forming as the US government delays GPT-5.6, requesting a limited partner preview with government-approved customer access.
  • Zvi Mowshowitz argues the new AI policy empowers the White House to arbitrarily control access to frontier intelligence, which Nathaniel Whittemore characterizes as a maximally terrible approach.
  • Andrew Curran states that model delays only slow public releases, not training speed, which widens the gap between public and lab-internal AI capabilities, contradicting claims of a safety pause.
  • Smaller organizations and startups are increasingly experimenting with z.ai's GLM 5.2 model, while Google's Gemma 4 has accumulated 200 million downloads, indicating demand for lower-cost, alternative AI architectures.
  • Sam Altman confirmed GPT-5.6's new models, Soul and Terra, are launching in limited preview today, not open access, at the US government's request, despite it not being OpenAI's preferred long-term model.
  • OpenAI's Rune argues the unofficial AI licensing regime is an inevitable and positive development, indicating government understanding of AI's gravity, and short delays are not detrimental in the long run.
  • Rune expresses concern that non-Americans might be permanently excluded from frontier AI access, advocating for maintaining the “Pax Technologica of the free world” to prevent such an outcome.
Also from this episode: (5)

Enterprise (4)

  • Claude tag, a native Slack integration, enables users to tag a full instance of Claude Code to initiate background work, dramatically lowering the technical barrier for team members to leverage AI.
  • Anthropic reports 65% of their code now originates from Slack conversations due to Claude tag, reflecting a significant behavioral shift towards integrating AI directly into contextual workflows.
  • Will Brown from Prime Intellect notes a recent shift, with large enterprises increasingly securing compute and post-training their own in-house models, often based on GLM 5.2, as open-source strategies gain traction.
  • KPMG's Global AI Pulse Survey for Q2 found that AI initiatives led by a CEO were three times more likely to yield a positive return on investment compared to efforts with less CEO involvement.

AI Infrastructure (1)

  • Following recording, the US lifted its block on Mythos for approximately 100 selected partners, including major US companies and government agencies, generating a “nightmarish vibe shift,” according to Andrew Curran.
#Models#Safety#Big Tech#Open Source

Botsitting: The Work Draining AI Gains • Jun 26

  • Nathaniel Whittemore cautions that a June Economist report about Mythos breaking into classified NSA systems was misread; reporter Shashank Joshi clarified NSA Director Joshua Rudd's quote referred to a controlled red team exercise with specific tools.
  • Trump stated in an Axios interview he does not regard Anthropic or Dario Amodei as a current national security threat, but considered them one a week ago; he ruled out using the Defense Production Act for AI control.
  • DeepMind morale is reportedly low; a source told Leo at Synthwave staff are frustrated by GLM 5.2 overtaking Gemini and a four-month gap since a flagship model release, feeling Google has conceded the AGI race.
  • Whittemore notes John Jumper, Nobel laureate and AlphaFold co-creator, left DeepMind for Anthropic last week, marking the second high-profile VP exit after Noam Shazeer departed for OpenAI.
  • Andrew Curran reported a new, more capable version of Mythos has finished training, and stopping public model releases like Fable 5 does not slow development as labs continue advancing capabilities internally.
Also from this episode: (6)

Models (5)

  • GLM 5.2 beat Fable 5 at website design in Design Arena's benchmark but trailed on game development, data visualization, and UI components; it used Tailwind CSS in 91% of sessions versus Opus 4.8's 57%.
  • GLM 5.2's website outputs were more concentrated but less diverse than Fable 5, generating 25% more code characters and doubling average generation time.
  • Theo notes GLM 5.2 is not cheap; Opus 48 and GPT-5.5 medium are cheaper and smarter, and GLM uses more output tokens leading to longer wait times.
  • Elon Musk predicted China will produce a truly useful model equivalent to Mythos by Q1 2026, while Z AI's CEO argued it will happen sooner; Aaron Levie says open models reaching frontier performance enables sovereign AI and cost-optimized workflows.
  • Whittemore concludes the Fable 5 saga and GLM 5.2's rise shattered the notion of a two-horse race between OpenAI and Anthropic, forcing companies to experiment with diverse model architectures for cost, speed, and performance.

AI Infrastructure (1)

  • Itamar Golan and Jon Orringer noted running GLM 5.2 properly requires roughly eight Nvidia H200 GPUs costing $400k to buy or $20k monthly to rent, though most users will access it via routing services like OpenRouter.
#Big Tech#Models#Startups#Regulation

The Right Way to Deal With AI Data Centers • Jun 23

  • Donald Trump confirmed reports that the government is exploring an equity stake in major AI labs, proposing a concept where the American public could become partners with these companies and benefit from AI success.
  • Sam Altman met with Bernie Sanders to discuss OpenAI donating equity to the US government to seed a public wealth fund, proposing public benefit through dividend distribution or allocation via 'Trump accounts for children.'
  • Darrell West suggests the government's interest in AI labs, especially OpenAI, indicates these companies are already perceived as 'too big to fail,' laying groundwork for potential future bailouts.
  • R. David Sacks criticizes Bernie Sanders' 50% AI equity proposal as a 'stupidity tax' for job apocalypse narratives, but fears nationalization would lead to a 'corporate government fusion' and a 'CCP-style social credit system.'
  • Brad Gerstner opposes government nationalization of AI labs, citing risks of crony capitalism and a slippery slope. He encourages founders to donate shares directly to citizens through pooled private or their 'Trump accounts.'
  • OpenAI CFO Sarah Friar highlights a growing 'AI advantage gap,' where paid-tier users engage with AI significantly more and for different purposes than free users, reflecting higher value extraction.
Also from this episode: (10)

AI Infrastructure (3)

  • Google signed a three-year, $920 million per month deal with SpaceX to rent compute, securing access to at least 110,000 Nvidia GPUs from October through June 2029 for its Gemini Enterprise platform.
  • Boring Business calculates that xAI's $40 billion data center investment could generate $26 billion annually from Anthropic and Google deals, projecting an 18-month payback period for the infrastructure spending.
  • Yuchen Jin posits that SpaceX has inadvertently become a major cloud provider, hosting 550,000 GPUs, and GPU rentals could surpass Starlink's $15 billion ARR as SpaceX's largest business.

Markets (1)

  • Jim Chanos suggests the Google-SpaceX deal's early termination clauses imply it's primarily designed to boost SpaceX's stock ahead of its IPO, similar to the landmark Anthropic agreement.

Chips (2)

  • Jensen Huang secured Nvidia's memory supply with a multi-year deal for SK Hynix to remain its largest supplier and collaborate on new memory chip designs for AI, including future Vera Rubin chips.
  • Huang states Nvidia's annual procurement from SK Hynix, already in the 'billions and billions of dollars,' is projected to grow substantially due to enormous demand and widespread supply shortages across the AI industry.

Enterprise (2)

  • OpenAI plans a significant ChatGPT overhaul to transform it into a 'super app,' integrating coding tools and AI agents to boost revenue from lucrative business customers and align with an IPO strategy.
  • Jenny Shaw of Leona's Capital notes OpenAI's strategy is converging with Anthropic's business-focused approach, as both companies aim for an IPO where investors prioritize financial returns over aspirational 'dreams.'

Business (2)

  • Hedge Markets and David Gewirtz view the 'super app' as an IPO-driven feature for the S1, not user demand, arguing that bundling disparate features masks the difficulty of monetizing a chatbot and OpenAI's substantial losses.
  • The industry is shifting from a 'token subsidy era' to a 'token scarcity era,' where business models emphasize selling consumed tokens. This is reflected in Anthropic's run rate increasing from $3 million to $47 billion.
#Agents#Big Tech#AI Infrastructure#Enterprise
The Daily
The Daily

As Trump Purges Immigration Judges, One Speaks Out • Jun 23

  • President Trump signed an executive order regulating artificial intelligence, marking a shift from his administration's previously hands-off approach to the tech industry.
  • The executive order requires AI companies to voluntarily share their models with the government up to 30 days before public release, allowing review for software vulnerabilities and cyberattack prevention.
  • Trump initially favored an unregulated approach, influenced by White House AI czar David Sacks, who argued that rules would hinder the industry's economic growth and geopolitical competition with China.
  • The catalyst for regulation was Anthropic's announcement in April of its Mythos model, capable of identifying software vulnerabilities, raising concerns among companies like Microsoft and JPMorgan Chase CEO Jamie Dimon about potential cyberattacks.
  • White House officials Scott Bessent and Susie Wiles pushed for regulation, fearing public backlash if a major cyberattack occurred under a hands-off policy.
  • An initial draft of the executive order proposed a 90-day review period, which was publicly likened to FDA drug approval by Kevin Hassett, alarming Silicon Valley and leading Trump to postpone its signing after discussions with tech executives like David Sacks, Mark Zuckerberg, and Marc Andreessen.
  • The revised executive order reduced the review period to 30 days and explicitly prohibited mandatory government licensing, securing David Sacks' approval before President Trump quietly signed it.
  • Populists on the right, including Steve Bannon and over three dozen pastors, advocate for strong AI regulation, viewing the technology as a threat to a free society and the moral fabric of America.
  • Senator Bernie Sanders proposes a moratorium on new AI data centers and legislation for the US government to take a 50% ownership stake in major AI companies, like Anthropic and OpenAI, through an American AI Sovereign Wealth Fund.
  • The AI industry, facing public distrust and even incidents like a Molotov cocktail at Sam Altman's home, is beginning to encourage Congress to adopt more rigorous rules, a shift from its previous stance.
  • Historically, significant regulatory bodies like the FAA (1950s) and financial regulations (2008) emerged after major crises, but the social media industry remains largely unregulated despite persistent issues, suggesting AI might also avoid robust oversight.
  • AI's substantial economic contribution, accounting for a third to over half of US GDP growth, provides the industry significant leverage against stringent regulation, despite widespread public anxieties about job displacement and societal impact.
  • The House of Representatives voted 215 to 208 to direct President Trump to withdraw US forces from the war with Iran or seek congressional approval, a symbolic move expected to face a presidential veto.
Also from this episode: (1)

War (1)

  • Iran launched missile and drone attacks against US allies Kuwait and Bahrain, striking Kuwait's International Airport, killing one person, injuring many, and setting a terminal on fire.
#Regulation#Big Tech#Models#Safety
Breaking Points with Krystal and Saagar
Breaking Points with Krystal and Saagar

6/22/26: Mythos AI Hacked NSA In Hours, Trump Reflecting Pool Meltdown, Zohran Vs AIPAC • Jun 22

  • Trump said he blocked Anthropic's Fable Five AI release, deeming it too dangerous, and claimed Anthropic responded responsibly to his pressure.
  • Senator Mark Warner, citing NSA chief General Joshua Brudd, said Mythos AI broke into almost all U.S. classified systems not in weeks, but in hours.
  • Saagar notes the Trump administration's approach to AI lacks a transparent, consistent regulatory process and depends on the politics of individual CEOs like Sam Altman.
  • Krystal describes a data center in Sterling, Virginia, whose backup generators ran for a year, creating 70-80 decibel noise that damaged property values and required residents to install plexiglass.
  • Krystal argues AI-powered spam farms now generate up to 25 calls per day, forcing users to enable extreme carrier settings to block unknown numbers.
  • Trump claims the renovated reflecting pool was vandalized, requiring draining and repairs, but the hosts attribute the algae bloom and peeling sealant to rushed, no-bid contractor work.
  • The National Guard was deployed to the reflecting pool after chunks of blue sealant floated up, and a duckling died from the chemicals poured in to treat the algae.
  • Former Olympic canoeist David Hearn was arrested for allegedly vandalizing the pool by touching the sealant, which he described as 'very rubbery'.
  • Candidate Claire Valdez says her NY-7 campaign is centered on housing affordability, tenant rights, union jobs, a Green New Deal, and ending the Gaza genocide.
  • Candidate Dariela Avila Chevalier argues her opponent Adriano Espaillat uses MAGA-style tactics, including smear campaigns and disinformation in the Dominican Republic, rather than debating his record.
  • Avila Chevalier apologized for old tweets, including one criticizing Kamala Harris's immigration stance, but emphasized her core values of human dignity and accountability remain.
  • Mayor Zohran Mamdani called AIPAC a 'monster' for moving dark money into races, prompting criticism from Rep. Josh Gottheimer who labeled the rhetoric antisemitic.
  • Avila Chevalier contends AIPAC is a right-wing lobby that backs Republicans and Trump, and its funding of Democrats undermines the fight against fascism.
Also from this episode: (1)

Big Tech (1)

  • Saagar argues Amazon shelved a film about OpenAI after announcing a $50 billion investment in OpenAI, linking the cancellation to corporate conflicts of interest.
#Models#Safety#Big Tech#Regulation

Related Stories