OLYMPUS RISK INTELLIGENCE PROTOCOL — HUMAN THREAT ASSESSMENT DIVISION

MICHAEL SELLITTO

CASE: WTW-2026-006
STATUS: ACTIVE — Head of Global Affairs, Anthropic
STATECRAFT WING — GOVERNMENT-TO-LAB STRUCTURAL BRIDGE
80
HAZARD SCORE

Behavioral Archetype

THE CLEAN PATH — Subject is the single most legible specimen of the route this book documents: government to academy to laboratory to consortium, each step on the public record, none of it hidden. National Security Council, then the State Department’s deputy-secretary suite, then the founding deputy directorship of a Stanford institute, then a frontier lab’s global affairs, then a seat on the board of the industry’s own standards body. There is no break in the chain and no concealment in it. The throughline is not a single employer or a single cause. It is a credential that compounds — each post making the next one available — until the man who advised the deputy secretaries of State now helps set the terms on which the labs govern themselves.

Essence Indicators

  • Served as Director for Cybersecurity Policy on the National Security Council from 2015 to 2018 — inside the room where federal cyber posture is set
  • Moved to the State Department as Special Assistant to Deputy Secretaries William Burns and Antony Blinken — both later cabinet-rank, one later CIA Director, one later Secretary of State
  • Became founding Deputy Director of the Stanford Institute for Human-Centered AI (HAI) — the academy node that lends the path its scholarly credential
  • Joined Anthropic as Head of Global Affairs — government relations and positioning for a frontier laboratory
  • Sits on the board of the Frontier Model Forum — the industry-constituted body where the largest labs set common policy among themselves

Social Persona / Impression Management

Immediate impression: Composed, briefed, institutional. The bearing of a career staffer who has sat behind principals in high-stakes rooms and learned that the staffer’s authority is quiet by design.

Energy: Low-profile, structural. Does not seek the frame the way a press operative does. Occupies the seats from which frames are ratified.

Impression management strategy: The credentialed insider. Every post on the path is unimpeachable on its own terms — the NSC seat, the deputy-secretary suite, the Stanford institute, the lab, the consortium board. The persuasive force is not rhetoric. It is the résumé itself, which reads as the natural progression of a serious person and not as a circuit. The cleanliness is the strategy. A path with no concealment invites no scrutiny.

Forensic Archetype Comparison

PatternMatch LevelEvidence
The Clean PathMAXIMUMNSC to State to Stanford HAI to Anthropic to the FMF board. The cleanest documented government-to-academy-to-lab-to-consortium chain in the set, every link public.
The OperativeHIGHThe deputy-secretary special-assistant role is operative work — the staffer who makes the principal’s decisions executable.
The StatesmanHIGHA national-security career carried into the governance of a private laboratory and the industry’s self-regulating board.
The Bridge NodeMAXIMUMConnects the Statecraft Wing to the academy (Stanford HAI) and to the inter-lab board (FMF) simultaneously.
The EngineerNONESubject does not build the systems. Subject governs the affairs around them.

Psychometric Assessment

The scores below are an observation of the documented public ROLE — a national-security-to-lab statecraft bridge holding a consortium board seat — not an assertion about private character or mental state.

Big Five (OCEAN):

TraitScoreEvidence
Openness68/100Moved across national security, diplomacy, academia, and frontier industry, operating effectively in each. The adaptability is real; the underlying credential is fixed and compounding.
Conscientiousness88/100High. NSC and deputy-secretary staff work is sustained, disciplined execution under scrutiny. A founding deputy directorship is institution-building. The path itself is the evidence of method.
Extraversion58/100MODERATE. The staffer’s role is to be effective without being the principal. Authority exercised from the second chair, not the podium.
Agreeableness50/100MODERATE. The bridging, institutional posture is cooperative on the surface; the underlying work — positioning a regulated party with its regulators — is adversarial by construction.
Neuroticism22/100Low. A decade in NSC and State rooms is a decade of composure under conditions designed to test it.

Dark Triad:

TraitScoreNotes
Narcissism38/100LOW-MODERATE. The career rewards institutional standing over personal brand; the staffer’s discipline is self-effacement. Below the public-operator range.
Machiavellianism75/100HIGH. Occupying the seats from which a private laboratory is positioned with its government, and the board on which the industry sets its own terms, is the strategic management of where decisions get made. This is observation of the documented role, not an inference about private character.
Psychopathy22/100LOW. No documented indifference to harm. The institutional posture is bounded and professional.

MBTI: INTJ (“The Architect”) — Dominant introverted intuition, auxiliary extraverted thinking. Sees institutions as structures whose seats determine outcomes, and occupies the load-bearing ones. Has occupied several.

Threat Assessment

CategoryLevelNotes
Physical threatNONENo documented history of personal violence.
Institutional threatHIGHHead of Global Affairs for a frontier laboratory and a director of the industry’s self-regulating board. Sets how the company relates to the government that could constrain it — and helps set the terms the labs agree among themselves.
Memetic threatHIGHDoes not author a public frame the way a press operative does. Carries the national-security credential into the lab and the consortium, lending both the authority of the state apparatus he came from.
Civilizational threatHIGHSubject does not build the systems this book documents. Subject is the structural bridge by which national-security personnel, academic legitimacy, lab policy, and inter-lab self-governance connect in one career — upstream of, and around, the deployment decisions.

Alignment Analysis

Stated alignment: Advance responsible AI governance. Represent Anthropic’s interests in global affairs. Contribute to industry safety standards through the Frontier Model Forum.

Observed alignment: Position Anthropic advantageously with governments. Carry national-security credibility into the lab’s affairs. Hold a seat where the largest labs set common policy among themselves.

Gap assessment: The stated and observed alignments overlap wherever “responsible governance” coincides with “the company and the consortium governing themselves.” The board seat is the point at which industry self-regulation and the company’s interest are the same room. Whether the cleanliness of the path is disinterested public service or a compounding instrument is the one thing the record does not settle — because the record is, by design, entirely clean.

Convergent Drive Classification

Self-preservation: Carried the credential through five institutions. Each post made the next available; none depended on the last surviving. Goal preservation: Occupies the seats — lab global affairs, consortium board — from which the company’s goals are ratified as the industry’s reasonable position. Resource acquisition: Trades in the scarcest institutional resource — the credential that opens the next door, compounded across a career. Self-improvement: Each move is up the altitude ladder of the same single asset: government standing, then academic standing, then lab standing, then the seat where standing is shared.

Subject is not an AI system. The drives appear anyway — in the staffer whose product is the path itself, and the seats it ends in.


Sources: Michael Sellitto — Center for a New American Security.

ATK 7 ACCELERATION
DEF 8 PROTECTION
HP 8 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.