KATE STARBIRD

Behavioral Archetype

THE ADVISOR ON BOTH SIDES OF THE SEAM — Subject is the academic partner in the Election Integrity Partnership who, two years later, chaired the federal advisory subcommittee that told CISA how to handle mis-, dis- and malinformation. The research and the advice are each defensible on their own terms. Together they place one researcher at both ends of the question the apparatus never settled: who decides what the government should do about speech it considers harmful. This profile scores that position, not the woman.

Essence Indicators

Professor of Human Centered Design & Engineering at the University of Washington; BS in Computer Science, Stanford (1997); PhD, University of Colorado (2012). Her research began in crisis informatics and, in her department’s words, “followed the phenomenon of online rumoring down the rabbit hole”; it holds that disinformation “is inherently participatory,” and her recent work concerns “right-wing media as effectively leveraging partisan, participatory dynamics through improvisational performances” (UW HCDE)

Co-founder of the UW Center for an Informed Public (2019), one of its five co-founders, and its faculty director for a two-year term from late 2021 (UW CIP; UW HCDE)

Election Integrity Partnership partner (2020). The CIP was one of the EIP’s four core partners with the Stanford Internet Observatory, Graphika and the Atlantic Council’s DFRLab; see Election Integrity Partnership and The Censorship-Industrial Complex

Chaired CISA’s MDM subcommittee (2022). As chair of the Cybersecurity Advisory Committee’s “Protecting Critical Infrastructure from Mis- Dis- and Mal (MDM) information” subcommittee she presented its June 22, 2022 recommendation that CISA “focus on addressing MDM risks that undermine critical functions of American society” and “invest in external research to assess the impact of MDM threats and the efficacy of its MDM mitigation efforts” (CISA readout)

Her side of the record. The CIP’s March 16, 2023 rebuttal: “CISA did not found, fund, or otherwise control the EIP. CISA did not send content to the EIP to analyze, and the EIP did not flag content to social media platforms on behalf of CISA”; the CISA collaboration ran July to November 2020 under a Trump appointee and was “reviewed and approved by Trump Administration attorneys” (UW CIP)

Hines v. Stamos. Subject was a named defendant in America First Legal’s suit over the EIP and the Virality Project (W.D. La. No. 3:23-cv-571, filed May 2, 2023). On July 1, 2026 Chief Judge Terry A. Doughty dismissed it without prejudice for lack of jurisdiction, finding that “discovery has confirmed the absence of any factual basis to conclude the EIP or VP, let alone any particular Defendant, coerced any social media platform to censor Plaintiffs’ posts or accounts”; the plaintiffs appealed (5th Cir. No. 26-30471) (ruling; judgment)

Immediate impression: The careful empiricist. Methods-first, dataset-first; describes rumoring as a crowd phenomenon rather than a villain’s plot.

Energy: Defensive under fire and specific about it. Answers charges point by point, in published rebuttals with dates.

Impression management strategy: The researcher who studies, never the operator who acts. The rebuttal is precise and much of it is documented. The position it does not address is the second chair: the same researcher advising the agency on what to do about the category of speech her partnership had flagged.

Forensic Archetype Comparison

Pattern	Match Level	Evidence
The Empiricist	HIGH	A decade of crisis-informatics and rumor research; the PECASE and SIGCHI Societal Impact awards her department lists.
The Advisor	HIGH	Chaired CISA’s MDM subcommittee and presented its recommendations in June 2022.
The Partner	HIGH	The CIP was a core EIP partner in 2020.
The Operator	CONTESTED	Critics place the EIP’s referrals on her; her center says platforms decided and CISA controlled nothing; the Hines court found no factual basis for coercion.

Psychometric Assessment

Big Five (OCEAN):

Trait	Score	Evidence
Openness	78/100	High. Moved from crisis informatics to political disinformation, and frames both as collective behaviour.
Conscientiousness	82/100	High. Dated, itemized public rebuttals; long-running research program.
Extraversion	48/100	MODERATE. Public when the work requires it; not a media personality.
Agreeableness	55/100	MODERATE. Collaborative in the lab; combative only in rebuttal.
Neuroticism	38/100	LOW-MODERATE. Kept publishing through suits, subpoenas and threats.

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

Trait	Score	Notes
Narcissism	30/100	LOW. The work is framed as the team’s and the field’s.
Machiavellianism	45/100	MODERATE. The dual role, partner to the flagging project and adviser to the agency, is structural leverage over what the government treats as a threat. Observation of the documented position, not of character.
Psychopathy	10/100	LOW. No documented indifference to harm.

MBTI: INTJ-adjacent, inferred from the documented public role only: systems-minded, data-first, precise in defense.

Threat Assessment

Category	Level	Notes
Physical threat	NONE	No documented history of personal violence.
Institutional threat	MODERATE-HIGH	Advised the federal agency at the centre of the dispute on its MDM role, while co-leading a centre that partnered in the 2020 flagging work.
Memetic threat	MODERATE	Her “participatory disinformation” frame is widely used in the field.
Civilizational threat	MODERATE	The hazard is the advisory seat, not the person: a researcher shaping how the state treats a speech category she studies.

Alignment Analysis

Stated alignment: Understand how rumors and disinformation spread, and “strengthen democratic discourse by building resilience to online misinformation, propaganda, and manipulation.”

Observed alignment: Partnered in the 2020 EIP; chaired the 2022 federal subcommittee advising CISA on MDM; defended both in detail.

Gap assessment: The record does not show the coercion her critics alleged; a federal court that looked after discovery found no factual basis for it. What the record does show is a researcher on both sides of the seam between study and policy. That needs no motive to be the finding.

Convergent Drive Classification

Self-preservation: Answers each charge in writing and in court, and kept the center running. Goal preservation: Holds the “participatory disinformation” frame as the field’s way of seeing the problem. Resource acquisition: Federal research funding and a federal advisory seat; the CIP’s institutional standing. Self-improvement: From rumor-tracking in disasters to advising a national agency on a speech category.

Subject is not an AI system. The drives appear anyway — in the advisor who studied the thing she was asked to advise on.

Sources: Kate Starbird — UW Human Centered Design & Engineering; Kate Starbird discusses CIP’s priorities as faculty director — UW CIP; Readout of CISA’s third Cybersecurity Advisory Committee meeting — CISA; UW CIP on Election Integrity Partnership research claims — UW CIP; Hines v. Stamos, Memorandum Ruling, Doc. 222 — W.D. La.; Hines v. Stamos, Judgment, Doc. 223 — W.D. La..

ATK 5 ACCELERATION

DEF 8 PROTECTION

HP 7 RESILIENCE

OLYMPUS RISK INTELLIGENCE PROTOCOL does not exist. It was assembled in a GitHub issue thread in October 2023 by engineers who had read the extinction risk letter and wanted to understand who specifically had signed a document saying AI might kill everyone and then continued working on AI. These dossiers are satire. The biographical facts cited are sourced from published reporting, public statements, academic papers, and court records. The psychometric scores are not clinical assessments. No part of this constitutes professional psychological evaluation or diagnosis. Do not use these dossiers to make decisions about anything.