GRACE WERNER

Behavioral Archetype

THE CO-FOUNDER WHO MOVED TO THE POLICY SHOP AND KEPT THE BOARD SEAT — Subject is a co-founder of AVERI, the evaluator that advocates mandatory third-party audits of frontier AI developers, who now runs the growth function of the Foundation for American Innovation, the Washington think tank whose supporters include the labs the audits would cover, and who still sits on AVERI’s board. AVERI’s team page says she “co-founded AVERI and served as Chief of Staff and lead U.S. policy”; FAI’s page says she “led strategy and U.S. policy research at AVERI” before becoming its Vice President of Growth. Her own site says: “I am the VP of Growth at the Foundation for American Innovation. Previously, I was the chief of staff and U.S. policy lead at AVERI.” The hazard is not a gate. She approves, funds and blocks nothing in the record. It is that the person who started the audit advocate now holds the fundraising and expansion portfolio at a think tank that lists Anthropic and OpenAI among its corporate supporters, while keeping the board seat. This profile scores the position, not the person.

Essence Indicators

• Board member and co-founder, AVERI. AVERI’s team page, under “Board”: “Grace Werner is a board member at AVERI and Vice President of Growth at the Foundation for American Innovation (FAI). Previously, Grace co-founded AVERI and served as Chief of Staff and lead U.S. policy. Before joining AVERI, Grace worked in corporate strategy and national security technology program development.” AVERI’s about page: “Our mission is to make frontier AI auditing effective and universal,” and “we do not accept cash donations or compensation from frontier AI companies or their executives”; it names among its funders “the AI Underwriting Company, Coefficient Giving, Craig Falls, Geoff Ralston, Good Forever Foundation, Fathom, Halcyon Futures, Sympatico Ventures” (AVERI, Team; AVERI, About). See AVERI and, on the same board, Max Henderson and Mike McCormick.

• Vice President of Growth, Foundation for American Innovation. FAI’s page: “Grace Werner is vice president of growth at the Foundation for American Innovation. Previously, she led strategy and U.S. policy research at AVERI; worked in global strategy at Visa; and managed projects on intelligence, surveillance, and reconnaissance at Sandia National Laboratories. Grace holds an MSc in international relations from the London School of Economics and a BA in political science from the University of New Mexico.” FAI’s Q1 2026 update says: “This quarter we welcomed Vice President of Growth Grace Werner.” FAI’s about page: “Our major philanthropic supporters have included Arnold Ventures, … Coefficient Giving, … Survival and Flourishing Fund, … the William and Flora Hewlett Foundation, and many others,” and “Our corporate supporters and event sponsors have included a16z, Alphabet, Amazon, Anthropic, … Meta, Microsoft, … OpenAI, …” (FAI, Grace Werner; FAI, Quarterly Update Q1 2026; FAI, About). See Foundation for American Innovation.

• Her own page. gracemwerner.com: “I am the VP of Growth at the Foundation for American Innovation. Previously, I was the chief of staff and U.S. policy lead at AVERI; researched the geopolitical dimensions of AI development; worked in global strategy at Visa; and managed projects on intelligence, surveillance, and reconnaissance at Sandia National Laboratories.” It lists “M.Sc. in international relations, international political economy, London School of Economics; B.A. in political science, summa cum laude, University of New Mexico,” a bookshelf, and an email address at thefai.org. It links no social account (Grace M. Werner).

• The signed quarterly letter (1 July 2026). FAI’s “Quarterly Update, Q2 2026,” bylined “Grace Werner”: “FAI is getting ready for our American Innovation Gala on October 1 in Washington, DC. … Keynote speakers include Shyam Sankar, Chief Technology Officer and Executive Vice President of Palantir Technologies, and inventor, author, and futurist Ray Kurzweil. Tickets and sponsorships are available; we hope you can join us.” It reports: “FAI’s experts represented one of three organizations working with Anthropic on a review of model behaviors around freedom of expression, including political conversations, as part of Anthropic’s work on election safeguards.” And: “We recently launched Frontier Legal Defense. This new project will defend America’s leading edge in artificial intelligence and other transformative emerging technologies from government overreach and novel legal threats while preserving alignment with the national interest.” It also says “Senior Fellow Dean Ball is joining OpenAI as Head of Strategic Futures. Dean will stay on with FAI as a non-resident senior fellow” (FAI, Quarterly Update, Q2 2026, 1 Jul 2026). See Dean Ball.

• Author, ‘Frontier AI Auditing’ (arXiv, 16 January 2026). She is the thirteenth of 48 listed authors, with AVERI’s executive director Miles Brundage first and, among others, Patricia Paskov (fifth), Conrad Stosz (sixth), Chris Painter, Dean Ball and Bri Treece. The paper’s note: “inclusion as an author does not entail endorsement of all claims in the paper, nor does authorship imply an endorsement on the part of any individual’s organization.” The paper defines frontier AI auditing as “rigorous third-party verification of frontier AI developers’ safety and security claims” (arXiv:2601.11699; arXiv HTML, v1). See Patricia Paskov.

• Author, AEF-1 (4 December 2025). The AI Evaluator Forum’s AEF-1, “Minimum Operating Conditions for Independent Third Party AI Evaluations,” lists “Grace Werner” fifteenth of 26 authors with affiliation “AI Verification and Evaluation Research Institute,” beside Stosz (first, Transluce), Brundage (AVERI), Paskov (Oxford Martin AI Governance Initiative) and others. “Version 1, updated December 4, 2025” (AI Evaluator Forum, AEF-1). See AI Evaluator Forum.

• Organizer, DC AI Security Forum (9 July 2025). The forum’s page, in the National Press Club, “Organized by Hamza Chaudry, Grace Werner, and Caleb Parikh,” lists “Grace Werner / Chief of Staff / Averi” among its named participants, and says it “is a project of Cambridge Boston Alignment Initiative, a nonprofit organization that is tax exempt under Section 501(c)(3).” Its stated core themes include “cyber capability evaluation” and “AI model weight protection” (AI Security Forum, DC AI Security Forum ‘25).

• Social accounts: none verified. Her own page, FAI’s profile of her and AVERI’s team page link no personal X account; FAI’s page links the organization’s. LinkedIn is not used as a source. No handle is added.

Social Persona / Impression Management

Immediate impression: The staff-builder. A strategy and national-security program background (Visa, Sandia), the chief of staff who co-founded the audit advocate, then the growth officer at a think tank. The footprint is two team pages, her own site, one signed letter, two author blocks and an event listing.

Energy: Operational. Her one signed text is a quarterly donor-and-friends letter that announces a gala, a project launch, hires and a research roundup; the record contains no op-ed, interview or talk by her on audits.

Impression management strategy: Credentials-first and list-shaped. Her page reads “I am the VP of Growth … Previously, I was the chief of staff and U.S. policy lead at AVERI; researched the geopolitical dimensions of AI development; worked in global strategy at Visa” (Grace M. Werner). The quarterly letter’s frame is FAI’s: “defend America’s leading edge” and “strengthen national security, improve American governance, and keep America at the forefront of technological dynamism” (FAI, Quarterly Update, Q2 2026).

First-hand exhibits. The signed Q2 2026 letter is the only extended text in her name; the rest is bio.

Forensic Archetype Comparison

Psychometric Assessment

Trait scores are observations of the documented public role, not claims about private character.

Big Five (OCEAN) (inferred from the public role and public record only):

Dark Triad (evidence-bound):

MBTI: ENTJ-adjacent (“The Commander”), inferred from the documented public role only. The rationale is the work product: a co-founder who then runs growth and writes the letter that opens a gala. First-hand rationale: “Previously, Grace co-founded AVERI” (AVERI, Team). A style inference, not a test result.

Threat Assessment

Score rationale. The receipts: AVERI’s team and about pages, FAI’s profile of her, about page and quarterly updates, her own page, the arXiv paper, the AEF-1 PDF and the DC AI Security Forum page. Attack 4, defense 3, durability 4, hazard 36.7. Peers: Ben Goldhaber (4/3/3, hazard 33.3), a program officer who places a funder’s money: level on attack and defense, a point below on durability, because she co-founded an institution and he authored none. Kai Fronsdal (4/3/4, hazard 36.7): level on all three, the author of an instrument where she is the co-founder of an advocate and a co-author of standards, the same persistence and the same dependence on organizations she does not run. Conrad Stosz (5/5/5, hazard 40), who chairs the Forum: a point or two above on each because the Forum’s page names him Chair, and her seat on a board with no named office is the lower rung. Dean Ball (6/5/6, hazard 56.7), her FAI colleague, who drafted the private-auditor proposal and now leads a policy team at a lab: two points above on attack and defense because his position places him inside a lab and hers does not. Attack 4 is the number most open to challenge: it rests on seats and authorship, and FAI’s page states a role in an Anthropic review, not an effect on any decision.

Score basis. ATK 4 is leverage: a board seat and co-founder’s role at AVERI, the growth portfolio at FAI, authorship of AEF-1 and the auditing paper, no gate (AVERI, Team; AEF-1). DEF 3 is insulation: a staff post and a board seat with no tenure, at organizations whose pages list their supporters (FAI, About). HP 4 is durability: a co-founded institution, two author blocks and a signed letter (arXiv).

Alignment Analysis

Stated alignment: Third-party audit of frontier AI developers, made rigorous. AVERI: “Our mission is to make frontier AI auditing effective and universal” (AVERI, About); and FAI’s: “keep America at the forefront of technological dynamism” (FAI, Quarterly Update Q2 2026). Observed alignment: She co-founded the audit advocate, wrote into its standard, then moved to the think tank whose supporters include the labs, keeping the board seat, and signs the letter that reports the think tank’s work with Anthropic and a legal-defence project against “government overreach” (AVERI, Team; FAI, Quarterly Update Q2 2026). Gap assessment: None documented. AVERI says it “discloses and carefully manages potential conflicts of interest” and publishes them on its Team page; her Team entry prints her two roles and no conflict statement (AVERI, About; AVERI, Team). What the record cannot settle is the dates of her AVERI roles (neither organization prints them) and whether the FAI role has been disclosed as a conflict in AVERI’s terms.

Interest: Documented, never motive, never a verdict. She has published no policy argument beyond co-authored technical prose and a quarterly letter; the positions below are the ones her documents carry.

1. Position: third-party evaluators should meet minimum operating conditions of access, independence and transparency (AEF-1: “it can be challenging to understand whether an evaluator had sufficient access to assess the characteristics of interest, whether they enjoyed meaningful independence, and how transparently they shared their methods and findings,” AEF-1). What adoption pays: AVERI, her affiliation on the document, is a founding member of the Forum that issued it, and its executive director writes that “some AVERI staff helped write their first standard, AEF-1” (Miles Brundage, The Launch of AVERI); she is its board member. No fee, grant or contract value for her is published. Best case: the standard’s own: independent evaluations are “becoming increasingly central to AI governance.” Others who gain: the eight member evaluators and the funders behind them.
2. Position: AI policy should defend American frontier developers against government overreach, with industry and philanthropy at the table (the letter she signs: “defend America’s leading edge in artificial intelligence and other transformative emerging technologies from government overreach and novel legal threats,” FAI, Quarterly Update Q2 2026). What adoption pays: FAI’s about page lists Anthropic, OpenAI, Meta, Microsoft and Alphabet among corporate supporters and event sponsors, and Coefficient Giving and the Survival and Flourishing Fund among philanthropic ones; her title is growth (FAI, About). Best case: FAI’s: “We help policymakers become more effective stewards of innovation.” Others who gain: the labs the project would shield, and the donors who underwrite FAI’s research.

Alliance Map

Several of these run at the same time. In 2026 she is on the board of the audit advocate, runs growth at the think tank that works with and is supported by the labs, is a co-author of the Forum’s standard, and wrote the donor letter that announces a gala keynoted by a defense-technology executive. The overlap is the finding, not a coordination claim. See the small world.

Convergent Drive Classification

Self-preservation: Moves with the work while keeping the seat. First-hand: “Previously, Grace co-founded AVERI” and “is a board member at AVERI” in one entry (AVERI, Team). Goal preservation: Keeps one object, the audit regime, in two organizations. First-hand: “make frontier AI auditing effective and universal” (AVERI, About). Resource acquisition: Her portfolio is growth: sponsorships, a gala, a larger team and office. First-hand: “A bigger team needs a bigger office” (FAI, Quarterly Update Q2 2026). Self-improvement: From national-security program management to an audit advocate’s chief of staff to a vice presidency. First-hand: “managed projects on intelligence, surveillance, and reconnaissance at Sandia National Laboratories” (Grace M. Werner).

Subject is not an AI system. The drives appear anyway — in the co-founder who kept the board seat and took the growth desk.

Public footprint: gracemwerner.com · FAI profile · AVERI team.

Sources: AVERI, Team; AVERI, About; FAI, Grace Werner; FAI, About; FAI, Quarterly Update Q1 2026; FAI, Quarterly Update Q2 2026; Grace M. Werner; arXiv:2601.11699, Frontier AI Auditing; arXiv HTML v1; AI Evaluator Forum, AEF-1; AI Security Forum, DC AI Security Forum ‘25; Miles Brundage, The Launch of AVERI.