ELHAM TABASSI

Behavioral Archetype

THE STANDARD-WRITER — Subject wrote the framework. The NIST AI Risk Management Framework — the voluntary document that supplies the United States its common vocabulary for what “trustworthy AI” means and how to manage its risk — was developed under her leadership, and Brookings calls her its lead author. The reach is not a deployment vote or a funding seat. It is that when a US company, agency, or contract needs a defensible answer to “how did you manage the AI risk,” the answer is increasingly we followed the NIST framework — the one she architected. She then carried that authority into the US AI Safety Institute as its first Chief Technology Officer, the federal body later renamed CAISI (cross-reference: us-caisi.md, subject #55). A voluntary standard that everyone adopts is law without a vote, and she holds the pen on the American one.

Essence Indicators

• Led the development of the NIST AI Risk Management Framework (AI RMF 1.0), released January 26, 2023; NIST’s own award page credits “her leadership in developing” it, and Brookings describes her as its “lead author”

• The AI RMF is voluntary, “intended for voluntary use to improve the ability to incorporate trustworthiness considerations into the design, development, use, and evaluation of AI products, services, and systems,” and its development was congressionally mandated under the FY2021 National Defense Authorization Act

• Long NIST career: joined in 1999, with roles including Electronics Engineer, Senior Research Scientist, Chief of Staff of the Information Technology Laboratory, and Associate Director for Emerging Technologies — over 25 years of civil service

• Earlier NIST work was in biometrics and image quality: a primary author of NFIQ 2 (NIST Fingerprint Image Quality 2), NIST Internal Report NIST.IR.8382, whose features were standardized in ISO/IEC 29794-4 and deployed by the FBI and DHS

• On February 7, 2024, named Chief Technology Officer of the US AI Safety Institute (Elizabeth Kelly named Director), responsible for shaping NIST and community efforts to “conduct research, develop guidance, and conduct evaluations of AI models including advanced large language models”; the USAISI was renamed the Center for AI Standards and Innovation (CAISI) in June 2025

• Named to the inaugural TIME100 Most Influential People in AI (announced September 7, 2023)

• Education: undergraduate degree in electrical engineering from Sharif University of Technology (Tehran); MSc in electrical and electronics engineering from Santa Clara University

Social Persona / Impression Management

Immediate impression: The career standards scientist. Precise, methodical, fluent in the neutral-technical register of a national metrology bureau. Reads as a measurement engineer, not a policymaker or an executive — which is the point.

Energy: Framework-first, consensus-through-process. Does not argue a model’s safety case; builds the standard that tells everyone how to argue their own, and runs the public comment process that legitimizes it.

Impression management strategy: The neutral standard. The most defensible posture in the federal evaluation layer: a NIST framework is, by institutional design, voluntary, consensus-built, and apolitical. The neutrality is genuine in form — and that is what makes it adoptable across an entire economy. The choice of which risks the framework foregrounds, and what “trustworthy” is defined to mean, is a choice, and the pen that made it was hers.

Forensic Archetype Comparison

Psychometric Assessment

Big Five (OCEAN):

Dark Triad:

MBTI: ISTJ (“The Logistician”) — Dominant introverted sensing, auxiliary extraverted thinking. Builds the rigorous, standardized, defensible procedure and maintains it. Treats AI risk the way she treated fingerprint-image quality: define the measurement, write the standard, make it the reference everyone cites.

Threat Assessment

Alignment Analysis

Stated alignment: Help organizations manage AI risk and incorporate trustworthiness into AI systems. Advance responsible, standards-based AI through a voluntary, consensus framework.

Observed alignment: Consistent. The framework exists, is voluntary, was built through public process, and is widely adopted; the USAISI CTO role applied the same standards expertise federally. The work is substantiated by the artifacts.

Gap assessment: No meaningful gap between stated and observed at the personal level — the file is reach-not-malice. The structural tension is the one the org-level file names (us-caisi.md, conduct: STATE-INSTRUMENT — REPOINTED FROM SAFETY TO RIVAL-GRADING BY ELECTION): the federal body she helped lead kept its desk and its access to frontier models while its stated purpose was rewritten by a change of administration, not by evidence. A “voluntary” framework that becomes the de facto compliance baseline is law without a vote — defensible when the pen is a career metrologist, instructive when the institution holding the pen can be repointed by election. Her individual alignment is clean; the leverage the framework confers is the finding, and it is structural.

Convergent Drive Classification

Subject is not an AI system, and exhibits none of the convergent drives. The relevant pattern is upstream of every American compliance claim: she wrote the framework by which AI risk is defined, managed, and reported. The convergent drives belong to the systems the framework governs; her reach is over the standard that decides what managing them responsibly even means. A voluntary framework, once the economy adopts it, acquires its own durability — it persists because rewriting it would unsettle everyone’s compliance posture — and its own quiet goal: the definition of “trustworthy” it encodes becomes the default everyone reasons inside. Subject held the pen. The reach is that an entire economy agreed to manage its AI risk in her framework’s terms.

Sources: NIST AI RMF 1.0 Launch — NIST; TIME Magazine recognizes NIST’s Elham Tabassi — NIST; Elham Tabassi — Brookings Institution; US AI Safety Institute leadership (Kelly, Tabassi) — FedScoop; US AI Safety Institute renamed CAISI — SSTI.