MICHAEL KRATSIOS

Behavioral Archetype

THE DOOR THAT REVOLVED THROUGH THE WHITE HOUSE TWICE — Subject entered government from the personal office of a Silicon Valley financier, ran the federal government’s technology-policy shop as deputy and then as confirmed chief, left for a company that sells AI systems to the Pentagon, and returned to the same category of federal post — this time with the authority to write the government’s national AI framework. The pattern is not a single revolving door. It is the same door, on the same hinge, opened in the same order twice: private capital, public office, private AI vendor, public office again, each pass at higher formal authority than the last. This profile scores the commute, not the man. The facts are the sequence: who held which chair, in what order, and what authority the last one carries.

Essence Indicators

• Served as principal at Thiel Capital and chief of staff to Peter Thiel before entering government — the venture-financier office from which he was recruited into the first Trump administration’s technology staff

• Appointed Deputy Chief Technology Officer of the United States in March 2017, reported at the time as the sole staff member in the White House CTO’s office

• Nominated Chief Technology Officer of the United States in March 2019, confirmed by the Senate by unanimous consent that August, and served through January 20, 2021

• Left government in 2021 to become Managing Director and Head of Strategy at Scale AI, a data-labeling and AI-evaluation company whose own government-facing marketing names the White House, US Army, US Air Force, CDAO, and Defense Innovation Unit as customers and advertises products cleared for classified networks

• Nominated Director of the White House Office of Science and Technology Policy in December 2024; confirmed by the Senate 74–25 on March 25, 2025, as the 13th person to hold the post

• As OSTP Director, is the named federal official quoted releasing “Winning the AI Race: America’s AI Action Plan” on July 23, 2025 — the document setting federal AI deployment, procurement, and export policy

• The same AI Action Plan process drew a formal comment from OpenAI — authored by Chris Lehane — framing the policy choice as “democratic AI” versus “autocratic AI,” meaning the office Kratsios now runs is also the one the labs it will govern lobby directly

• Spoke for the United States at the UN Security Council on 23 September 2026, the 10228th meeting, the session at which Yoshua Bengio, Sam Altman, Dario Amodei and Clément Delangue briefed. His words, in order: “the most consequential technological development of our time, superintelligence, formerly known as AI”; “The frontier of intelligence is advancing rapidly. That is not a reason to pause its further development or to constrain it with new global governance structures”; “international dialogue in this forum and in others cannot be allowed to drift towards global governance. As President Trump said before the General Assembly yesterday, the United States totally rejects any attempt to construct a globalist scheme of control of superintelligence”; “The American people’s representatives will legislate and regulate on the American people’s behalf. You should do the same for your people”; “You cannot govern technology you do not understand”; and “A prosperous future will not be secured by a global regulator. It will be secured by sovereign nations that adopt superintelligence, responsible companies that build it, and free people who refuse to be ruled by fear.” He told the Council the US had “engaged frontier labs on testing and evaluation of new model capabilities, established cybersecurity clearinghouses for coordination on vulnerability mitigation, and clarified the existing law enforcement tools available to address harmful uses of superintelligence,” and cited the G20 “Carolina principles” on preserving “national sovereignty in the governance of emerging technologies” (UN transcript, S/PV.10228)

• The new vocabulary he used at the Council became an order six days later: Executive Order 14434, “Inaugurating the Era of Super Intelligence” (29 September 2026), orders the executive branch to “use the terms ‘Super Intelligence’ and ‘SI’ in place of ‘Artificial Intelligence’ and ‘AI’,” applies that to agency “websites” among other documents, and gives the Assistant to the President for Science and Technology 60 days to submit “proposed legislative language to establish a Federal definition of ‘Super Intelligence’.” Within days NIST’s CAISI page reappeared as the “Center for Advancing Innovation and Standards for Super Intelligence (CAISSI)” (nist.gov/caissi)

• His own questionnaire to the Senate Commerce Committee lists the sequence in his words: Associate, then Chief Compliance Officer, then Chief Financial Officer at Clarium Capital Management (2010 to 2015); Chief Compliance Officer at Thiel Capital and Thiel Macro (2012 to 2014); “Principal and Chief of Staff, Thiel Capital, 2014-2017”; Deputy U.S. Chief Technology Officer (2017 to 2019); Chief Technology Officer of the United States and Associate Director of OSTP (2019 to 2021); Acting Under Secretary of Defense for Research and Engineering (2020 to 2021); Managing Director, Scale AI (2021 to 2025). He lists the Defense Department’s Medal for Distinguished Public Service among his recognitions (Senate Hearing 119-219, Kratsios questionnaire)

• Told the Commerce Committee on 25 February 2025 what the office was for: “As President Trump has said, ‘as our global competitors race to exploit these technologies, it is a national security imperative for the United States to achieve and maintain unquestioned and unchallenged global technological dominance.’” He added that “The shape of future global order will be defined by whomever leads across AI, quantum, nuclear, and other critical and emerging technologies,” and that the way to lead was “the uniquely American free-market approach to scientific discovery, which harnesses the combined strengths of government, industry, and academia” (Senate Hearing 119-219, prepared statement). Asked about an AI action plan, he put the federal measurement agency at the centre of it: “NIST has a long bipartisan history of being an agency that does measurement science,” and the work on any AI initiative should be in that light (same hearing)

• Disclosed to the same committee that “I hold equity and options in Scale AI, bitcoin, and ethereum. I will divest of these assets if confirmed, per my ethics agreement.” Asked by Ranking Member Cantwell, in writing, how he had ensured Scale’s contract data workers were fairly compensated and protected from traumatic material, he answered each time that “While at Scale AI, I did not work on the part of the business” that dealt with those contractors, which Scale calls contributors (Senate Hearing 119-219, ethics questionnaire and responses to Senator Cantwell). Separately he answered the legal-matters question: “I am named as a defendant in a putative class action filed on December 10, 2024 against a former employer by individuals alleging California labor law claims related to their status as independent contractors”

• His 2026 annual financial disclosure (OGE Form 278e, appointment date 03/2025, signed 10 June 2026, certified by the agency ethics official and by the Office of Government Ethics) lists his Scale AI position as Managing Director from 3/2021 to 1/2025, shows Scale AI salary income of $21,953, and records that “I will continue to participate in my former employer’s 401(k) plan and HSA” for Scale and in the Thiel Capital 401(k); its asset schedule lists no Scale AI equity (OGE Form 278e, Kratsios, 2026 annual report)

• The White House’s own director page calls him “President Donald J. Trump’s chief science and technology advisor” and says that in the first administration he “served as the fourth Chief Technology Officer of the United States at the White House” and as Under Secretary of War for Research and Engineering at the Pentagon; it lists only the White House’s institutional accounts, no personal one. In the present tense it says he is “the thirteenth Director of the White House Office of Science and Technology Policy” and leads the administration’s efforts on “critical and emerging technologies such as super intelligence, quantum computing, and biotechnology” (The Director, whitehouse.gov/ostp)

• Is named in the orders that set the federal AI agenda. Executive Order 14179 (January 2025) tasks “the Assistant to the President for Science and Technology (APST),” with the Special Advisor for AI and Crypto and the National Security Advisor, to develop the AI Action Plan within 180 days, and states the policy “to sustain and enhance America’s global AI dominance” (EO 14179). Executive Order 14363, “Launching the Genesis Mission” (24 November 2025), says the APST “shall provide general leadership of the Mission” (EO 14363). Executive Order 14365, “Ensuring a National Policy Framework for Artificial Intelligence” (11 December 2025), says the APST and the Special Advisor for AI and Crypto “shall jointly prepare a legislative recommendation establishing a uniform Federal policy framework for AI” (EO 14365). The Super Intelligence order of 29 September 2026 gives the same office the 60-day task already cited above

• Signed the foreword to the White House’s year-one science and technology report, which says the Genesis Mission “is the most ambitious scientific undertaking since the Apollo program,” and lists “Preventing a Patchwork of AI Regulations” among the AI wins: “A single national framework will decrease barriers to entry in AI and promote innovation, while ensuring American companies can compete on the world stage and are not subject to restrictions that their international competitors do not face.” It quotes him on the Genesis order: “The Genesis Mission connects world-class scientific data with the most advanced American AI to unlock breakthroughs in medicine, energy, materials science, and beyond” (24 November 2025) (Trump Administration Science & Technology Highlights: Year One, OSTP, January 2026)

• Co-chairs the President’s Council of Advisors on Science and Technology with David Sacks: “The Council will be co-chaired by David Sacks and Michael Kratsios.” The 25 March 2026 announcement names thirteen appointees: Marc Andreessen, Sergey Brin, Safra Catz, Michael Dell, Jacob DeWitte, Fred Ehrsam, Larry Ellison, David Friedberg, Jensen Huang, John Martinis, Bob Mumgaard, Lisa Su and Mark Zuckerberg (The White House, 25 Mar 2026). See Marc Andreessen and David Sacks

• On 21 July 2026 released “Science: A New Golden Age,” a report to the President modelled on Vannevar Bush’s postwar report, with an FY2028 research-and-development budget memo issued jointly with the OMB Director. His statement: “American scientific progress was the beating heart of the 20th century.” The report calls to “fully fund and scale the Genesis Mission as the flagship national AI-for-science initiative” and to “diversify funding mechanisms beyond slow consensus peer review,” and agencies must submit an action plan within 90 days (The White House, 21 Jul 2026)

• Hosted the G20 Innovation Ministerial with the Department of Commerce at Chapel Hill, which closed on 2 September 2026 with a consensus statement and “Carolina Principles”; he said “flexible policy frameworks crafted to promote innovation will drive economic growth and prosperity” (The White House, 2 Sep 2026). The UN Security Council statement three weeks later cited those principles

On the Record

His own words as the White House and the Senate publish them, in date order. The Security Council statement of 23 September 2026 is quoted at length in the Essence Indicators above.

The hearing, 25 February 2025. The opening statement: “I drafted non-regulatory guidance for emerging technologies to give the public trust in how these technologies are developed and used,” of his first-term work, and on China, “The shape of future global order will be defined by whomever leads across AI, quantum, nuclear, and other critical and emerging technologies.” On Scale, in writing: “While at Scale AI, I did not work on the part of the business” that dealt with its contract workers (Senate Hearing 119-219).

The Action Plan, 23 July 2025. “America’s AI Action Plan charts a decisive course to cement U.S. dominance in artificial intelligence” (The White House, 23 Jul 2025).

The Genesis Mission, 24 November 2025. “The Genesis Mission connects world-class scientific data with the most advanced American AI to unlock breakthroughs in medicine, energy, materials science, and beyond” (OSTP, Year One).

The year-one foreword, January 2026. The administration is “aligning policy, regulatory frameworks, and funding authorities to accelerate mission-driven research and development,” and “The American AI Exports Program is bringing together diplomatic and development resources to support the American technology stack all over the world” (OSTP, Year One).

The science report, 21 July 2026. “American scientific progress was the beating heart of the 20th century” (The White House).

The G20 ministerial, 2 September 2026. “flexible policy frameworks crafted to promote innovation will drive economic growth and prosperity” (The White House).

The Security Council, 23 September 2026. “You cannot govern technology you do not understand” (UN transcript, S/PV.10228).

His X account. The White House director page links only the White House’s institutional accounts, so no personal X handle is attached to this page (whitehouse.gov/ostp/the-director).

Social Persona / Impression Management

Immediate impression: The interchangeable technocrat. Young for the résumé, unshowy, photographed in the same dark suit across two administrations and one company. Reads as competent process management rather than as a public advocate for any single position.

Energy: Institutional, not rhetorical. Does not argue that a particular model is safe or dangerous. Occupies the chair that decides what “the government’s AI policy” means this year.

Impression management strategy: The reliable inheritor. The office existed before him and will exist after him; the framing is that whoever holds the confirmed technology-policy chair should be someone the last administration also trusted with a version of the same job. That framing is not false — both confirmations passed with bipartisan Senate support. What the record adds is that the chapter between the two federal tours was spent running strategy at a company whose paying customers are the same government departments that the federal AI framework now governs.

Three first-hand exhibits show the method. At the nomination hearing his opening statement hands the thesis to someone else, “As President Trump has said,” and attaches the goal of “unquestioned and unchallenged global technological dominance” to the President’s name rather than his own (Senate Hearing 119-219). On the labor questions about his former employer he gives the same one-line answer three times and does not argue the premises (same record). And the year-one report he signs names the policy “Preventing a Patchwork of AI Regulations” and describes it as protecting “American AI innovation from an inconsistent and costly compliance regime of 50 different state laws,” which is a frame stated as administration, not as advocacy (OSTP, Year One).

Social accounts and feeds: no personal account and no feed. His director page links only the White House’s institutional accounts, @whitehouse on X among them (The Director, whitehouse.gov/ostp).

Forensic Archetype Comparison

Psychometric Assessment

Big Five (OCEAN):

Dark Triad (held low and evidence-bound; the score measures structural position, not character):

MBTI: ESTJ-adjacent — process-first, institution-first. Reads as someone who inherits an existing office and runs it to greater scope rather than one who authors a new doctrine from outside. Exhibit: his stated method at the hearing is inheritance and coordination, “advising the President and coordinating the national initiatives” (Senate Hearing 119-219).

Threat Assessment

Alignment Analysis

Stated alignment: Per the White House’s own release, “chart a decisive course to cement U.S. dominance in artificial intelligence” — remove regulatory barriers, accelerate American AI infrastructure and innovation, and lead internationally on AI diplomacy and security.

Observed alignment: Held the same category of federal technology-policy post twice, with an intervening executive tenure at a company selling AI-labeling and evaluation services to the Pentagon and allied agencies; now authors and publicly announces the document that sets national AI deployment and procurement policy, for a process the regulated labs themselves fed comments into.

Gap assessment: No gap is documented between what he says the office is for and what the office does; the hazard here is structural, not a documented instance of misconduct. When the person writing the federal government’s AI Action Plan spent the prior chapter as an executive at a company whose paying customers are the same departments the plan now governs, “government AI policy” and “the defense-AI vendor’s preferred policy environment” share a formative professional experience even absent any instruction passing between them. The public record establishes the sequence. It does not establish, and this file does not assert, coordination or private motive.

The first-hand record adds the shape of the stated side. His own hearing statement defines the mission as dominance (“unquestioned and unchallenged global technological dominance”), his Security Council statement locates governance in “sovereign nations” and “responsible companies,” and his year-one foreword counts a single national framework and an AI exports program among the wins (Senate Hearing 119-219; UN transcript, S/PV.10228; OSTP, Year One). The observed record matches the stated side at each point where it can be checked: the orders assign his office the work he describes. Where the two come apart is only in what the stated side leaves out. The Security Council statement says the US “engaged frontier labs on testing and evaluation,” and names no body, binding rule or published result, so what the testing is and who sets its terms is not on the record he published.

Interest: Documented, never motive, never a verdict.

1. Position: one Federal framework in place of state AI laws (the year-one report: “A single national framework will decrease barriers to entry in AI and promote innovation,” OSTP, Year One; EO 14365 has his office “jointly prepare a legislative recommendation establishing a uniform Federal policy framework for AI,” EO 14365). What adoption pays: a developer or vendor operating in many states faces one compliance regime instead of fifty; his own disclosures show no equity in any such company (Scale equity pledged for divestiture, none on the 2026 schedule), so the documented payer is the industry the rule covers, not him. His best case: the report’s own: the patchwork is “an inconsistent and costly compliance regime of 50 different state laws.” Others who gain: Google, whose Kent Walker filed in March 2025 asking Washington to “preempt a chaotic patchwork of state-level rules on frontier AI development”; the other frontier labs; and the co-preparer of the recommendation, David Sacks, whose profile records Craft Ventures holdings in AI companies. The convergence between a lab’s filing and the order that followed it is the finding.
2. Position: no global governance of superintelligence, sovereign regulation instead (“international dialogue in this forum and in others cannot be allowed to drift towards global governance”; “The American people’s representatives will legislate and regulate on the American people’s behalf. You should do the same for your people,” UN transcript, S/PV.10228). What adoption pays: no global regulator sits over American developers, and measurement stays with a national agency, NIST, which he described at his hearing as an agency that “does measurement science” (Senate Hearing 119-219). His best case: “You cannot govern technology you do not understand.” Others who gain: the American labs that briefed the Council, Sam Altman and Dario Amodei among them, and any government that prefers to keep its own regulation.
3. Position: export the American AI technology stack (“The American AI Exports Program is bringing together diplomatic and development resources to support the American technology stack all over the world,” OSTP, Year One). What adoption pays: federal diplomatic and development support goes to the sale of American chips, clouds and models abroad; the report names no product or vendor in that sentence. His best case: the hearing’s: “it is a national security imperative for the United States to achieve and maintain unquestioned and unchallenged global technological dominance.” Others who gain: the American vendors in each layer of the stack.
4. Position: fund the Genesis Mission as the flagship AI-for-science program (“fully fund and scale the Genesis Mission as the flagship national AI-for-science initiative,” The White House, 21 Jul 2026). What adoption pays: federal money and partnership terms for the Department of Energy laboratories and for the industry participants the report says signed research agreements with DOE on 18 December 2025, 24 in number; the page names none, and the office he leads has “general leadership” of the program (EO 14363). His best case: “The Genesis Mission connects world-class scientific data with the most advanced American AI to unlock breakthroughs in medicine, energy, materials science, and beyond.” Others who gain: the laboratories, the universities and the 24 participants.

Alliance Map

The Corporate and Agency rows overlap in time and that is the finding: in 2024 and to January 2025 he was a Scale AI managing director, a Dunes Point Capital adviser, a Stanford fellow and a Foundation for American Innovation board member at once, and from March 2025 the same person holds the federal seat that writes the AI Action Plan and, from March 2026, co-chairs a council whose appointees include Jensen Huang, Larry Ellison, Michael Dell, Lisa Su and Mark Zuckerberg. The Thiel Capital 401(k) and Scale AI 401(k) and HSA are the continuing ties his filing discloses.

Convergent Drive Classification

Self-preservation: Carries one operating method — occupy the federal science-and-technology chair — across two presidential administrations and one company tenure in between. First-hand: his questionnaire lists the same category of post in 2017, 2019, 2020 and 2025 (Senate Hearing 119-219). Goal preservation: Authors and announces the AI Action Plan that sets the terms against which federal AI policy, and by extension every agency and lab operating under it, is measured. First-hand: the year-one report he signs says the Action Plan “has provided a north star for the nation” (OSTP, Year One). Resource acquisition: Holds the scarcest resource in current AI policy — the Senate-confirmed federal chair whose signature document sets the frame other AI regulation, state and federal, is measured against. First-hand: since March 2026 he also co-chairs the President’s science council with the AI czar, and his July 2026 report asks to “fully fund and scale the Genesis Mission” (The White House, 25 Mar 2026; The White House, 21 Jul 2026). Self-improvement: Each move redeploys the same instrument at greater formal authority — deputy, then confirmed CTO, then confirmed OSTP Director with a national AI framework attached to the title. First-hand: the Senate record lists the 2017, 2019 and 2020 posts in his own hand, and the 2025 and 2026 orders add the Genesis Mission and the Super Intelligence definition to the office’s portfolio (Senate Hearing 119-219; EO 14434).

Subject is not an AI system. The drives appear anyway — in the federal chair that keeps turning out to be the same door, on the same hinge, opening onto more each time it swings back.

Sources: Michael Kratsios — Wikipedia; Senate roll-call vote on the Kratsios OSTP nomination; OSTP confirmation press release — The White House; White House unveils America’s AI Action Plan; Scale AI — Government; OpenAI response to the OSTP/NSF RFI on the AI Action Plan; UN Security Council, 10228th meeting — UN transcript; Executive Order 14434 — The White House; CAISSI — NIST; research: the Security Council, the lunch and the xAI question; Senate Hearing 119-219, nominations of Kratsios and Meador, 25 Feb 2025 (GPO); OGE Form 278e, Kratsios, 2026 annual report; The Director, whitehouse.gov/ostp; EO 14179; EO 14363, Launching the Genesis Mission; EO 14365; Trump Administration Science & Technology Highlights: Year One (OSTP); President’s Council of Advisors on Science and Technology appointments, 25 Mar 2026; OSTP Director releases “Science: A New Golden Age,” 21 Jul 2026; G20 Innovation Ministerial concludes, 2 Sep 2026; Kent Walker profile; Scale AI profile; Peter Thiel profile.