# David Sacks — News Log Rolling chronological record of news developments. Major items that change the structural analysis get promoted to [[David Sacks]]. This log preserves the full context and source links. For the donation history, see [[Sacks Political Donations Timeline]]. --- ## 2026-10-04 — The Czar Job Goes to a Name That Was Not on the List, and Sacks Lands Exactly Where He Wanted: Outside It **Source:** NPR, October 4, 2026; TechRepublic, October 6, 2026; Defense One, October 5, 2026. ### 🔴 Major **Trump gave the AI czar job to [[Jay Clayton]], the sitting Director of National Intelligence, who appeared on none of the five-name shortlist Axios published on September 21 — and [[David Sacks]] is listed in the founding membership of the new [[Super Intelligence Force]] as an "external adviser," alongside [[Condoleezza Rice]].** The 2026-09-21 entry recorded five candidates: [[Scott Kupor]], [[Sriram Krishnan]], [[Chamath Palihapitiya]], Sean Cairncross and [[Emil Michael]]. Two of them — Kupor and Michael — are now vice chairs of the body rather than its head. The chair went outside the list entirely, to a securities lawyer running the intelligence community. That is the cleanest possible confirmation of the structural read this log has been making: the czar title was never the asset. Look at what Sacks ended up with. He left the formal job in March 2026, became co-chair of the President's Council of Advisors on Science and Technology, and is now an unpaid-sounding "external adviser" to the task force that replaced his old office — with no confirmation, no disclosure obligation attached to the adviser slot in any reporting reviewed, no 120-day deliverable, and no accountability for whatever the Force concludes. Meanwhile the Force's argument is his argument. The [[Super Intelligence Force]] mandate as reported includes avoiding overregulation; the 2026-09-29 executive order it operationalizes does nothing but rename the field; and Trump's framing at the September 29 White House summit was that there "should be tremendous self-regulation." This log recorded on 2026-09-12–13 that Sacks called coordinated restraint a cartel and told Amodei and Altman to slow down on their own time. The position is now the government's position, staffed by two of his network's people, chaired by someone with no prior AI-policy record and therefore no competing doctrine to bring. One wrinkle worth keeping an eye on, because it cuts against the pattern. TechRepublic's account of the Sacks adviser slot sits oddly beside the WGCU/NPR characterization that, unlike Trump, Sacks "backed warnings about AI risks and supports development slowdowns" — a description that contradicts both his own September statements and everything in this log. **Treat that characterization as a reporting error until a second outlet repeats it;** it is single-sourced and inconsistent with the primary record. What is corroborated across NPR, Defense One and TechRepublic is the membership list, the 120-day clock, and the chair. **Sources:** - [NPR: "Trump names national intelligence chief Jay Clayton as new AI czar"](https://www.npr.org/2026/10/04/nx-s1-5990781/jay-clayton-ai-czar-trump) - [TechRepublic: "Trump Names Jay Clayton AI Czar to Lead New 'Super Intelligence Force'"](https://techrepublic.com/article/news-trump-jay-clayton-ai-czar-super-intelligence-force) - [Defense One: "US spy chief tapped as White House AI czar"](https://www.defenseone.com/policy/2026/10/us-intel-chief-white-house-ai-czar/416416/) - [WGCU/NPR: "Trump names national intelligence chief Jay Clayton as new AI czar"](https://www.wgcu.org/2026-10-04/trump-names-national-intelligence-chief-jay-clayton-as-new-ai-czar) --- ## 2026-09-21 — Five Names for the Czar Job, and Sacks Is Not One of Them **Source:** Axios, September 21, 2026. ⚠️ The candidate list and the attribution about Sacks's intentions are single-sourced to Axios. ### 🟡 Notable **Two days after [[Donald Trump]] announced an "AI Force" and a search for a new AI czar, Axios named five people under consideration: [[Scott Kupor]] (OPM director), [[Sriram Krishnan]] (former senior White House AI policy adviser), [[Chamath Palihapitiya]], Sean Cairncross (national cyber director) and [[Emil Michael]] (Pentagon CTO). On [[David Sacks]]: he left the role in March 2026, remains involved as an adviser, and per people close to him "likely won't take up the mantle again."** The 2026-09-19 entry recorded that Trump's announcement confirmed, in passing, that Sacks had not held the job since March. This closes the open question that entry left: the people around him are briefing that he does not want it back. Worth taking at face value — and worth reading structurally, because the candidate list is itself the argument for why he does not need it. Four of the five are inside the network Sacks convenes. Palihapitiya is his own [[All-In Podcast]] co-host. Krishnan wrote the AI policy Sacks was nominally coordinating. Kupor came out of [[Andreessen Horowitz]]. Michael is the Pentagon's chief technology officer. A czar drawn from that list is not a check on the position Sacks occupies in this policy area; it is a staffing of it. The formal office was always the less valuable of the two assets he held, and it is the one he gave up. The framing shift in the same reporting is the part with a longer half-life. Axios describes the administration moving the AI safety debate onto economic ground — the multitrillion-dollar buildout — and recasting advocates of a slowdown as "doomers." This log recorded on 2026-09-12–13 that Sacks told Amodei and Altman to slow down all they like and called coordinated restraint a cartel. That argument has now been adopted at the presidential level, with Trump saying he will "cherish" the technology and address harms "through the existing criminal justice system." The position outlived the appointment, which is the whole case for why leaving the job cost him nothing. **Sources:** - [Axios: "Trump wants an 'AI Force' to fight 'doomers' — and a scramble for AI czar is underway"](https://www.axios.com/2026/09/21/trump-ai-force-czar) - [Axios: "Trump wants a new AI czar and an 'AI Force' modeled on Space Force"](https://www.axios.com/2026/09/19/trump-ai-czar-space-force-safety) - [CNN: "Trump vows to create 'AI Force' and appoint czar amid calls to regulate technology's development"](https://www.cnn.com/2026/09/19/politics/trump-ai-task-force-czar) --- ## 2026-09-19 — Trump Announces an "AI Force" and a Search for a New AI Czar — Confirming, in Passing, That Sacks Has Not Held the Job Since March **Source:** Axios; Fox News; CNN; Washington Post, September 19, 2026. Truth Social post. ### 🔴 Structurally Significant **[[Donald Trump]] announced via [[Truth Social]] that he will create an "AI Force" modeled on the Space Force and will soon appoint a new AI czar — "Only High I.Q. individuals need apply!" He said the government would "not in any way hinder or stifle" the AI industry, would monitor only for "BAD" behavior through existing law, and dismissed AI safety concerns as a "hoax," arguing that a "strong and smart" president is the only safeguard required. No replacement name was given.** The reporting around the announcement carries a correction this log should absorb: multiple outlets describe [[David Sacks]] as the **former** White House AI and crypto czar, having stepped down from the formal role in March 2026 (TechCrunch, March 26), and now serving as co-chair of the President's Council of Advisors on Science and Technology. Axios describes him as still "influential" but no longer the czar. Separately, NBC News reported on September 18 that Sacks — again identified as "former" czar — will attend the state dinner for [[Xi Jinping]] on September 24. **Two things follow.** First, a housekeeping one: entries on this page since March have described Sacks as holding a post he had already left, and the substantive question going forward is whether his influence ran through the title or around it. The evidence points to around it — the September 12–13 "cartel" remarks, the open-weights reversal in August, and the [[JD Vance]] All-In appearance three days before this announcement were all made from a podcast chair, not an office. Second, and larger: the czar role is being refilled at the moment the administration is publicly repudiating the entire premise of AI safety oversight. "AI safety is a hoax" and "a strong and smart president is the safeguard" are not a deregulatory position, which would still imply rules applied lightly. They are a claim that the function belongs to one person's judgment. An "AI Force" on the Space Force model is a bureaucratic container for that claim — a uniformed service reporting up through the executive rather than a regulator with statutory authority. Whoever takes the czar title inherits an office whose job description was just narrowed from *policy* to *advocacy*, which may be exactly why the previous occupant is more useful outside it. **Sources:** - [Axios: "Trump wants a new AI czar and an 'AI Force' modeled on Space Force"](https://www.axios.com/2026/09/19/trump-ai-czar-space-force-safety) - [CNN: "Trump vows to create 'AI Force' and appoint czar amid calls to regulate technology's development"](https://www.cnn.com/2026/09/19/politics/trump-ai-task-force-czar) - [Washington Post: "Trump to form 'AI Force,' name AI czar but rejects calls for constraints"](https://www.washingtonpost.com/politics/2026/09/19/trump-form-ai-force-name-ai-czar-rejects-calls-constraints/) - [Fox News: "Trump announces new 'AI Force,' vows to protect industry as AI czar announcement nears"](https://www.foxnews.com/politics/trump-announces-new-ai-force-vows-protect-industry-ai-czar-announcement-nears) --- ## 2026-09-12–13 — "Go Ahead": Sacks Tells Amodei and Altman to Slow Down All They Like — and Calls Doing It Together a Cartel **Source:** Dario Amodei, "We Must Pace the Frontier," September 12, 2026; CDM, September 13, 2026; Truth on the Market, September 14, 2026; Townhall, September 14, 2026; Neowin. ### 🔴 Major **[[Dario Amodei]] published "We Must Pace the Frontier" on September 12, proposing employee-level access for third-party safety evaluators, coordinated safety standards among leading Western labs, and eventual coordination with governments including authoritarian ones. [[Sam Altman]] endorsed it within hours; [[Elon Musk]] followed minutes later ("Dario is right"). [[David Sacks]] answered that the proposal is an antitrust problem wearing a safety argument.** As quoted: *"Stop pretending you need to suspend antitrust law in order to create a cartel. Stop pretending METR is independent of [[Anthropic]] when they are literally intertwined with Anthropic's investors and staff. And stop pretending you need these same people to police your competitors who are not even at the frontier of AI."* A closely parallel rendering circulating elsewhere runs: *"Dario has written that we need to 'pace the frontier,' and Sam has agreed. People may be surprised by my response: go ahead. But stop pretending you need anyone else's permission..."* He characterized the two labs as a duopoly and the proposed framework as "regulatory capture" and "a DMV for AI." **Quote caution:** two slightly different renderings circulate across outlets and the original X post could not be retrieved. Attribute to the outlet rather than quoting as verbatim primary text. The **date** — September 12–13 — is solid. **Why notable.** This is the same move the August 20 entry below documents, aimed one step further in. There, the argument was that restricting open-weight models would kneecap American startups; the villain was safety regulation in the abstract. Here the villain is named — [[Anthropic]] and [[OpenAI]] specifically, with METR as the alleged instrument — and the legal frame is antitrust rather than competitiveness. The rhetorical structure is identical across both and across the June 19 Fable account and the July 3 adoption of Karp's broadside: whatever the frontier labs propose, the reason they propose it is to eliminate competitors, and the person making the argument is a [[President's Council of Advisors on Science and Technology]] co-chair whose venture portfolio depends on cheap model access. **The argument is also not obviously wrong, which is what makes it effective.** Coordinated pacing among the two or three firms at the frontier is, structurally, a horizontal agreement among dominant competitors to restrict output — the thing antitrust law exists to prohibit — and the fact that the stated purpose is safety does not make the structure disappear. Truth on the Market made essentially that case on September 14 from a conventional antitrust posture. What the KB should track is not whether Sacks is right but that the administration's ecosystem has now found an argument against AI safety coordination that does not require arguing safety is unimportant. Watch whether any actual antitrust authority picks it up. **Sources:** - [Dario Amodei: "We Must Pace the Frontier"](https://darioamodei.com/post/we-must-pace-the-frontier) - [CDM: "Amodei, Altman, and Musk all want to slow down AI — David Sacks says they are building a cartel"](https://cdm.press/news/business/2026/09/13/amodei-altman-and-musk-all-want-to-slow-down-ai-david-sacks-says-they-are-building-a-cartel/) - [Truth on the Market: "Move Slow and Collude: The Antitrust Problem With Pacing AI"](https://truthonthemarket.com/2026/09/14/move-slow-and-collude-the-antitrust-problem-with-pacing-ai/) - [Townhall: "David Sacks: Why do AI companies need government permission to slow down"](https://townhall.com/news/dmitri-bolt/2026/09/14/david-sacks-why-do-ai-companies-need-government-permission-to-slow-down-n2682937) - [Neowin: "David Sacks blasts OpenAI and Anthropic over slowdown"](https://www.neowin.net/news/david-sacks-blasts-openai-and-anthropic-over-slowdown/) --- ## 2026-08-20 — The Open-Weights Reversal: After a Legal-Tech Firm Post-Trains on Kimi K3, Sacks Argues Restricting Open Models Would Only Kneecap American Startups **Source:** Law.com, August 20, 2026; Harvey company blog, August 20, 2026; South China Morning Post, August 21, 2026; Benzinga, August 22, 2026 (Sacks post). ### 🟡 Notable **Harvey, the OpenAI-backed legal-technology company, announced on August 20 that its first in-house model, Tenet, was post-trained on top of Moonshot AI's open-weight Kimi K3 — a Chinese base model — and Sacks responded by publicly defending open weights, six weeks after treating the same model's benchmark performance as a national-security alarm.** The technical claim: Harvey took the 2.8-trillion-parameter Kimi K3 open-weight base, post-trained it on legal data with Fireworks AI, and reports state-of-the-art performance on legal benchmarks at a fraction of frontier-model cost. Sacks cited it as evidence that American companies build world-class specialized models on open bases, and argued that restrictions on open models would do nothing to stop Chinese labs from shipping the next Kimi while crippling startups like Harvey — adding that some closed labs would welcome such restrictions precisely because they eliminate competition. The reversal is what makes it worth logging. The July 17 entry below has Sacks reacting to Kimi K3's release by warning the U.S. could lose the AI race and blaming "regulatory capture" seekers. The threat object then was Chinese open-weight capability. The threat object now is American regulation of open weights, with Chinese open-weight capability recast as the raw material of American competitiveness. Both positions arrive at the same policy conclusion — no restrictions — from opposite premises about the same model. **Why notable:** This is the clearest example yet of the pattern this log has tracked since Sacks shed the czar title: the argument is downstream of the conclusion, and the conclusion is stable across contradictory facts. Note also who benefits from the specific framing. "Closed labs want restrictions because restrictions eliminate competition" is an argument aimed at [[Anthropic]] and, less directly, at [[OpenAI]] — the same targets as the June 19 Fable account and the July 3 adoption of Karp's broadside — from a PCAST co-chair whose own venture firm's portfolio depends on cheap model access. Watch whether this position survives contact with the first genuinely embarrassing Kimi-derived deployment. **Sourcing caveat:** the Harvey/Tenet launch is confirmed across Law.com, the company's own blog and SCMP. The Sacks quotes come from a single Benzinga writeup carrying an AI-assistance disclaimer plus his X post; the verbatim wording should be checked against the original post before being quoted elsewhere in the KB. **Sources:** - [Law.com: "Harvey Introduces Tenet, Its First Post-Trained AI Model for Legal"](https://www.law.com/legaltechnews/2026/08/20/harvey-introduces-tenet-its-first-post-trained-ai-model-for-legal-/) - [Harvey: "Post-training update — Harvey Tenet"](https://www.harvey.ai/blog/post-training-update-harvey-tenet) - [SCMP: "OpenAI-backed legal tech firm pivots to Chinese Kimi K3 open-weight model"](https://www.scmp.com/tech/tech-trends/article/3364827/openai-backed-legal-tech-firm-pivots-chinese-kimi-k3-open-weight-model) - [Benzinga: "David Sacks says restricting open models won't stop China"](https://www.benzinga.com/markets/tech/26/08/61371303/david-sacks-open-source-ai-harvey-tenet-model) --- ## 2026-08-14 — "Dark GPUs Would Be a Disaster": Sacks Names Overbuild as the Real Risk to Nvidia's $500B Financing Push — Then Argues the Anti-Data-Center Backlash Is What Prevents It **Source:** All-In Podcast, August 14, 2026; Wccftech, August 16, 2026; Barchart via Yahoo Finance, August 19, 2026. ### 🟡 Notable **Four days after [[NVIDIA]] announced memoranda with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR to mobilize "over $500 billion of third-party capital," Sacks named the failure mode: *"The biggest risk, to me, is not on the demand side, the biggest risk is that you get a glut of compute and you get an overbuild... if you have dark GPUs that would be a disaster for everyone."*** The dark-fiber analogy is the one the AI-infrastructure trade has been avoiding, and Sacks reached for it directly, tying the math to [[Elon Musk]]'s stated $30–$50 per watt compute valuation. Then he reversed: *"all of the political headwinds ensure against that outcome. Because it is so hard to build data centers... There is a whole moral panic slash hysteria slash hoax going on... those political headwinds I think will almost guarantee that there is not an oversupply relative to the exponentially growing demand."* Gavin Baker, on the same episode, called Nvidia "the central bank of AI, the Federal Reserve of AI." **Why notable:** The structure of the argument is the finding, not the forecast. Sacks concedes that the buildout's real risk is a capital glut, then treats the permitting fights, moratoriums, ratepayer revolts and state large-load tariffs tracked in [[The Quest for 90 Gigawatts]] as the mechanism that saves the industry from itself — while calling that same opposition a "hoax." That is a supply-constraint bull case dressed as a culture-war complaint, and it inverts the industry's public position, which is that the political headwinds are the threat. Barchart flagged the obvious conflict: Sacks co-chairs PCAST alongside Nvidia's chief executive, and was assessing the risk profile of a $500 billion Nvidia financing structure. **Sources:** - [Wccftech: "David Sacks calls the biggest risk to Nvidia's $500 billion AI financing plan a 'dark GPU' glut"](https://wccftech.com/us-tech-advisor-david-sacks-calls-the-biggest-risk-to-nvidias-500-billion-ai-financing-plan-a-dark-gpu-glut/) - [Barchart via Yahoo Finance: "David Sacks says biggest risk..."](https://finance.yahoo.com/technology/ai/articles/david-sacks-says-biggest-risk-164258160.html) --- ## 2026-08-11 — Craft Ventures V: Sacks Returns to Fundraising With a ~$1B Target, His First Flagship Since the Czar Job **Source:** The Information (Aug 11), via TechStartups; SEC Form D filed Aug 7. ### 🟡 Notable **[[Craft Ventures]] has formally launched Craft Ventures V, LP and is seeking roughly $1 billion — its first flagship fund since [[David Sacks]] left the White House AI-and-crypto czar role in late March, and a raise that would push the firm past $4 billion in total limited-partner capital.** The Form D, dated August 7, identifies the vehicle as a Delaware limited partnership operating from Craft's San Francisco address, lists the offering amount as indefinite, and reports no capital sold and no investors as of filing — an early-stage paperwork event, not a close. Sacks is listed as an executive officer through his role as managing member of the fund's ultimate general partner; Craft president Mark Woolway signed. The $1B figure comes from The Information's sources, not the filing, and remains subject to commitments. The escalation curve is the context: ~$350M for Fund I, $500M in 2019, $612M for Fund III plus a $510M growth fund in 2021, then a $712M Fund IV alongside a $608M second growth fund in 2023 — roughly $3.3B in AUM at that point. Craft kept investing throughout Sacks's Washington stint and opened an Austin office. Sacks remains co-chair of PCAST alongside Michael Kratsios. **Why notable:** This is the return leg of the revolving door, and the reporting says the quiet part out loud — The Information's sources framed Sacks's raised political profile as an *asset* for attracting limited partners and founders. The ethics record he leaves behind is what makes that legible: White House disclosures showed Sacks and Craft divested more than $200M in crypto-related and other positions before or around the start of his government service, a divestiture that is now sunk cost against a franchise he can re-raise at scale. Watch who the LPs turn out to be — sovereign and strategic money in a fund run by a sitting PCAST co-chair would be the conflict story, and Form D filings do not disclose it. Sources: - [TechStartups: "David Sacks' Craft Ventures targets $1 billion for first new fund since leaving White House"](https://techstartups.com/2026/08/11/david-sacks-craft-ventures-targets-1-billion-for-first-new-fund-since-leaving-white-house/) - [SEC Form D: Craft Ventures V, LP (filed Aug 7, 2026)](https://www.sec.gov/Archives/edgar/data/2140032/000214003226000001/0002140032-26-000001.txt) - [The Information: "Sacks' Craft Targets $1 Billion, First Fund Since White House Stint"](https://www.theinformation.com/articles/sacks-craft-targets-1-billion-first-fund-since-white-house-stint) --- ## 2026-07-17 — After Moonshot's Kimi K3 Breakout, Sacks Warns the U.S. Could Lose the AI Race — and Blames "Regulatory Capture" Seekers **Source:** Axios (July 17); The Hill (July 17). ### 🟡 Notable **From his PCAST co-chair perch, Sacks reacted to the release of Kimi K3 — an open-weights model from Chinese startup Moonshot AI that benchmarks on par with [[Anthropic]]'s and [[OpenAI]]'s frontier models — by warning the U.S. is in danger of losing the AI race, while insisting it will win if it holds to Trump's "pro-innovation approach" and rejects "bureaucratic controls fueled by hysteria, fear, and companies seeking regulatory capture."** The construction is the tell: the threat is Chinese open-source capability, but the villain of the framing is domestic — safety regulation and the (unnamed, but by now ritual) labs that supposedly seek it. It is the same rhetorical maneuver as the June 19 All-In account of the Anthropic break and the July 3 adoption of Karp's broadside, now stapled to a genuine competitive datapoint. Coverage noted Sacks retains real influence from PCAST despite shedding the czar title in March. **Why notable:** Kimi K3 hands the accelerationist wing its strongest evidence yet that the "China wins if we regulate" argument has stakes — and Sacks converted it within days into ammunition against the same regulatory apparatus his faction has been dismantling. Watch whether the administration cites Kimi K3 in rolling back compute-export or safety-testing rules, and how the frame holds up against the CAISI director's July 20 resignation. Sources: - [Axios: "David Sacks warns U.S. could lose AI race after Chinese model's breakout"](https://www.axios.com/2026/07/17/sacks-kimi-open-source-weights-trump) - [The Hill: "David Sacks warned US risks losing AI race amid new Chinese model launch"](https://thehill.com/policy/technology/5975303-david-sacks-us-artificial-intelligence-kimi-k3-moonshot-ai/) --- ## 2026-07-03 — Sacks (and Chamath) Adopt Karp's "Effing Insane" Broadside: "Real AI Safety" Recast as Enterprise Data Control, With Anthropic's Figma Move as Exhibit A **Source:** Web search sweep, July 5, 2026. ### 🟡 Notable **After [[Alex Karp]]'s July 1 CNBC eruption against the [[OpenAI]]/[[Anthropic]] token business model (delivered while fronting the [[Palantir]]–[[NVIDIA]] sovereign-AI deal), Sacks declared Karp "exactly right" to warn enterprises against feeding proprietary data to frontier labs — on X and then at length on the [[All-In Podcast]], where [[Chamath Palihapitiya]] joined the framing.** Sacks's X post dismissed "Legacy Media types" calling the Karp interview a "crash-out," and argued Karp "is articulating what real 'AI safety' looks like in the enterprise" — not alignment research or certification, but control over compute, models, data stack, and proprietary "alpha." As Exhibit A, Sacks pointed to Anthropic "blindsiding" Figma by launching the competing Claude Design tool after an Anthropic executive had served on Figma's board, part of a pattern (Claude Science, Claude Security, Claude Legal) of the lab expanding into categories served by companies building on its models. The move extends the 2026-06-19 pattern: the former czar functioning as the administration ecosystem's on-record narrator against the frontier labs — now explicitly redefining "AI safety" away from the alignment framing used by [[Anthropic]] and toward an enterprise-sovereignty framing that happens to align with Palantir's sales pitch and the administration's anti-Anthropic posture. Sources: - [Benzinga: "David Sacks Says Palantir's Alex Karp Was 'Exactly Right' to Warn Enterprises Against Feeding Data to OpenAI and Anthropic — the Figma Story Shows Why"](https://www.benzinga.com/markets/tech/26/07/60268293/david-sacks-says-palantirs-alex-karp-was-exactly-right-to-warn-enterprises-against-feeding-data-to-openai-and-anthropic-the-figma-story-shows-why) - [David Sacks on X: "Legacy Media types are calling this Alex Karp interview a 'crash-out'…"](https://x.com/DavidSacks/status/2072673187666813226) - [ABAB News: "Sacks and Chamath Support Karp's Warning: Companies Should Not Share Proprietary Data with OpenAI/Anthropic"](https://www.ababnews.com/news/4a510816-de9a-490c-b4a0-47b4015aa6d3) - [CNBC (context, July 1): "Palantir's Karp bashes OpenAI, Anthropic token model"](https://www.cnbc.com/2026/07/01/palantir-karp-open-ai-anthropic-tokens.html) --- ## 2026-06-19 — Sacks Gives the Insider Account of the White House–Anthropic Break on All-In, Casting Himself as Narrator of the Administration Line **Source:** Web search sweep, July 3, 2026. ### 🔴 Major **On [[All-In Podcast]] episode 277 (June 19), Sacks — speaking as former czar and PCAST co-chair — gave a behind-the-scenes account of the administration's clash with [[Anthropic]] over its "Fable" (Mythos-class) model, claiming private-sector testing partners flagged jailbreak vulnerabilities raising "cyber weapon" concerns, that the White House tried to reach Anthropic and "Dario refused," and that the breakdown in trust drove the government response, including the Pentagon's "supply chain risk" designation.** These are Sacks's characterizations, not confirmed by Anthropic or by official documentation; he amplified the same narrative in a long X thread framed as "what I believe to be true." The underlying clash predates this entry (Fortune reported June 14 on how a warning from Amazon led the White House to move against the model). What's new is the role: months after formally shedding the czar title, Sacks is functioning as the administration's on-record narrator and enforcer against a specific AI lab — influence-without-title in exactly the pattern the 2026-06-18 entry flagged, now aimed at shaping which frontier labs the government treats as trustworthy. Sources: - [All-In episode index: E277, "World's First Trillionaire, Anthropic Fable Banned, The New Oligarchs, Iran Peace Deal"](https://allin.onpodcastai.com/episodes) - [Benzinga: "David Sacks Says Anthropic Lost White House Trust Over Alleged AI Jailbreak Vulnerabilities"](https://www.benzinga.com/markets/tech/26/06/60004972/david-sacks-says-anthropic-lost-white-house-trust-over-alleged-ai-jailbreak-vulnerabilities-cyber-weapon-capabilities-what-this-comes-down-to-is) - [David Sacks on X: thread on the Anthropic situation](https://x.com/DavidSacks/status/2065853007619588171) - [Fortune (context, June 14): "How a warning from Amazon led the White House to shut down Anthropic's Mythos model"](https://fortune.com/2026/06/14/how-a-warning-from-amazon-led-the-white-house-to-shut-down-anthropics-mythos-model/) --- ## 2026-06-18 — Czar Title Dropped but Influence Migrates to PCAST; CLARITY Act Stalls; Craft Ventures Layoffs; Conflict-of-Interest Record **Source:** Web search sweep across November 2025 – June 2026. ### 🔴 Structurally Significant (promoted to main entry) **Sacks drops "AI/Crypto Czar" title, shifts to PCAST co-chair — influence preserved, scrutiny shed (March 25–26, 2026)** Sacks announced he had "used up" his 130 days as a special government employee and was transitioning out of the AI/Crypto Czar role into co-chairing the reconstituted [[President's Council of Advisors on Science and Technology]] alongside [[Michael Kratsios]]. The White House said it does not plan to appoint a replacement czar. Reporting across outlets was blunt that this was a relabeling, not a retreat: Axios framed it as Sacks dropping the "AI czar" label, "not policy influence." **Why structural:** This is the central move in the Sacks story. The czar title carried disclosure obligations and a hard 130-day clock; PCAST carries neither in the same form. The policies Sacks shaped — the [[GENIUS Act]] framework and the pending [[CLARITY Act]] — remain in force or in play while he keeps a seat at the table. PCAST's reconstituted roster ([[Marc Andreessen]], Fred Ehrsam, Sergey Brin, Michael Dell, Jensen Huang, Lisa Su, Mark Zuckerberg) institutionalizes the same tech-and-crypto network inside the government's senior science advisory body. The pattern echoes the [[Eric Schmidt News Log]] playbook: influence-through-advisory-perch rather than confirmed office. Sources: - [Axios: "David Sacks drops 'AI czar' label, not policy influence"](https://www.axios.com/2026/03/26/trump-white-house-david-sacks-ai-czar-policy-influence) - [CNBC: "David Sacks says his time as Trump's crypto and AI czar has ended"](https://www.cnbc.com/2026/03/26/david-sacks-trump-crypto-ai-czar.html) - [CoinDesk: "White House crypto czar David Sacks transfers to presidential advisory committee role"](https://www.coindesk.com/policy/2026/03/26/white-house-crypto-czar-david-sacks-transfers-to-presidential-advisory-committee-role) - [TechCrunch: "David Sacks is done as AI czar — here's what he's doing instead"](https://techcrunch.com/2026/03/26/david-sacks-is-done-as-ai-czar-heres-what-hes-doing-instead/) ### 🟡 Notable **CLARITY Act stalls in Senate as Sacks exits the czar role (March–June 2026)** Sacks left the czar post with his signature crypto legislation unfinished. The stablecoin-focused [[GENIUS Act]] had passed, but the market-structure [[CLARITY Act]] — which would split crypto oversight between the SEC and CFTC — cleared the House and then stalled in the Senate Banking Committee. Sacks had personally urged the crypto industry to back CLARITY in January and warned that [[Brian Armstrong]]'s withdrawal of support was a risk to the entire legislative agenda. A Tillis–Alsobrooks compromise on stablecoin yield cleared one roadblock, with a committee markup targeted for spring 2026. **Why notable:** The unfinished market-structure bill is the through-line connecting Sacks's government role to [[Craft Ventures]]'s crypto book and the broader [[Crypto-Political Machine]]. His continued advocacy from PCAST keeps a private-portfolio interest wired into an active legislative fight. Sources: - [Decrypt: "David Sacks Leaves White House Crypto Role With Key Legislation Still Unresolved"](https://decrypt.co/362531/david-sacks-leaves-white-house-crypto-role-legislation) - [The Block: "Crypto Czar David Sacks leaves post to go to broader White House tech role as legislation faces challenges"](https://www.theblock.co/post/395399/crypto-czar-david-sacks-leaves-post-to-go-to-broader-white-house-tech-role-as-legislation-faces-challenges) **Craft Ventures lays off staff, co-founder steps back, pivots to late-stage (early April 2026)** The Information reported that [[Craft Ventures]] laid off roughly half a dozen employees — primarily recruiters — and that a co-founder stepped back, as the firm refocused on late-stage investing. The restructuring came while Sacks was still nominally in government. **Why notable:** The firm Sacks built remained financially active and was reshaping its strategy during his government tenure — the backdrop against which the conflict-of-interest questions are read. Source: [The Information (via X): "Craft Ventures Lays Off Staff, Co-Founder Steps Back"](https://x.com/theinformation/status/1908889951275569184) **Conflict-of-interest record hardens: NYT, Warren/Stansbury, Public Citizen (Nov 2025 – 2026)** The ethics file that trailed Sacks throughout his czar tenure continued to build. A November 30, 2025 NYT investigation ("Silicon Valley's Man in the White House Is Benefiting Himself and His Friends") reported he held investments across 449 AI companies among 708 tech holdings. Warren and Stansbury launched a congressional investigation into his ethics waiver and possible breach of the 130-day SGE limit; Warren separately pressed the Office of Government Ethics ahead of a Trump crypto fundraiser. Public Citizen called for his resignation. Sacks's spokesperson called the conflicts narrative "false," saying government work had cost him money rather than made it. **Why notable:** These threads are already captured on the main [[David Sacks]] entry under Conflicts of Interest and feed [[Defense Tech Conflicts of Interest]]. Logged here for the running timeline; the structural point is that the unresolved ethics questions followed Sacks out of the czar role and into PCAST rather than being settled by his departure. Sources: - [NPR: "Trump tech adviser David Sacks under fire over vast AI investments"](https://www.npr.org/2025/12/12/nx-s1-5631823/david-sacks-ai-advisor-investment-conflicts) - [Warren: "Members of Congress Launch Investigation into Trump Crypto & AI Czar's Ethics Conflicts"](https://www.warren.senate.gov/newsroom/press-releases/warren-stansbury-members-of-congress-launch-investigation-into-trump-crypto-and-ai-czars-ethics-conflicts) - [Public Citizen: "David Sacks's Financial Conflicts… He Should Resign"](https://www.citizen.org/news/david-sacks-financial-conflicts-mean-he-can-no-longer-credibly-serve-as-white-house-ai-advisor-he-should-resign/) ### ⚪ Context / Color **All-In summit / sponsorship-access concerns.** Reporting noted that a planned White House AI summit was initially expected to be associated with the [[All-In Podcast]] Sacks co-hosts, with some staff worried sponsorship packages would create the appearance of selling access. No major new structural development beyond the existing podcast → fundraiser → policy pipeline already documented in [[All-In Podcast Defense Tech Pipeline]]. Source: [The Block: "White House AI and crypto czar David Sacks rejects fresh conflict of interest claims"](https://www.theblock.co/post/380916/white-house-ai-crypto-czar-david-sacks-rejects-conflict-of-interest-claims)