OLYMPUS RISK INTELLIGENCE PROTOCOL — INSTITUTIONAL ASSESSMENT DIVISION

GRAPHIKA

CASE: WTW-2026-051
STATUS: ACTIVE — Commercial network-analysis / influence-mapping firm, founded 2013
FRAMEWORK WING — INFLUENCE-MAPPING INSTRUMENT AUTHORITY
74
HAZARD SCORE — REACH
CONDUCT: CONFLICTED

OLYMPUS opened an institutional file. A firm has no Big Five and no Dark Triad, and the unit does not invent them; what a firm has is a mandate, a funding model, and a voice. Graphika is catalogued here as the cartographer — the commercial shop that draws the maps the rest of the field reasons about. The finding is the shape of the institution and who pays it: a private company whose network graphs became the documentary basis for public claims about who is influencing whom. Not a hand. A map, and the question of who commissioned it. The “breach reach” of an influence map is real, but it lives in prose; the numbers in the front matter are reach, funding entrenchment, and durability — not malice.

Institutional Archetype

THE MAP-MAKER — The archetype is the cartographer whose map becomes the territory everyone navigates by. Graphika does not run platforms and does not write anyone’s content rules; it analyzes the relationships between accounts and the flow of information across networks, and renders them as graphs. A map of who connects to whom, once adopted, structures what the reader believes is coordinated and what is organic — and the map-maker, not the reader, drew the boundary. That is the structural power: not authorship of any single takedown, but authorship of the picture inside which takedowns are argued. The instrument is the network graph. The leverage is that other institutions cite the graph as fact.

Mandate & Origin

Graphika was founded in 2013 by Dr. John Kelly (founder and Executive Chairman), a researcher in network and computational analysis of online behavior. It describes itself as providing “decision intelligence that helps organizations navigate complex digital environments,” using “network analysis, cross-platform intelligence, and expert-led research” to map how information moves and identify emerging risks. Its highest-profile public output came in December 2018, when Graphika — alongside CEO John Kelly and Oxford University researchers — produced one of the two reports commissioned by the U.S. Senate Select Committee on Intelligence analyzing Russia’s use of social media in the 2016 U.S. election. Camille François (cross-reference camille-francois.md) served as Graphika’s Chief Innovation Officer and authored the ABC (Actor–Behavior–Content) framework the field adopted; Ben Nimmo (cross-reference ben-nimmo.md) led cross-platform investigations there between DFRLab and Meta.

Funding & Backers

Graphika is a commercial firm, not a grant-funded lab — the distinction is load-bearing. Its self-described clientele spans “financial services, government, technology, and media,” and it has done work for Fortune 500 companies, platforms, and government bodies including the U.S. Senate Select Committee on Intelligence contract behind the 2018 report. The funding model is the finding: a private company whose maps are commissioned by — and sold to — the platforms and governments those same maps describe. Who pays for a map is part of what the map is.

Institutional Voice & Intent

The voice is the neutral-technical analyst register — the cartographer’s, not the advocate’s. Graphika speaks in the grammar of “decision intelligence,” “emerging risks,” and “complex digital environments”: measured, risk-management vocabulary that frames the firm as a navigation instrument rather than a participant. The persuasion is in the apparent objectivity of a graph — a network diagram reads as a finding, not an argument, even though someone chose its boundaries.

Stated intent: Help organizations understand complex digital environments early enough to make confident decisions before risks escalate; map influence operations rigorously and cross-platform.

Observed intent: Be the firm whose network maps the field, the platforms, and the government cite as the picture of who is coordinating with whom — the commercial instrument of record for influence attribution.

Gap: The stated and observed intents overlap wherever “map digital environments rigorously” coincides with “sell the authoritative map to the platforms and governments the map describes.” The SSCI commission is the clean example: a private firm’s network analysis became a foundational document for how the U.S. government and the public understood foreign influence — useful, sourced, and commissioned. Whether the commercial-map-of-record posture is neutral service or quiet positioning is not establishable from the outside, and for the map-maker it never needs to be. The recurrence — that the same firm’s people authored the field’s framework (ABC) and then carried it into a frontier lab and a platform — is the finding. The hand is not asserted.

Position in the Apparatus

Graphika is a hinge node. Its alumni populate the rest of this file: Camille François (Framework wing — author of ABC, now President of ROOST), Ben Nimmo (Connector spine — DFRLab → Graphika → Meta → OpenAI). The firm sits between the open-source-investigations world (DFRLab, where Nimmo came from) and the commercial and lab worlds (Meta, OpenAI, ROOST, where its people went). The revolving door is documented and lawful; the adjacency is recurrence, not a roster anyone curated. The map-maker’s graduates became the field’s framework authors and the labs’ threat-intelligence leads. No cabal. A circuit.

Actions & Leadership Choices

Founding purpose, judged on evidence. Graphika was founded in 2013 by Dr. John Kelly as a commercial network-analysis firm — and the commercial status is the founding fact, not a footnote. Judged on its deeds, its purpose is to sell the authoritative map of online influence to the parties with an interest in that map: its self-described clientele is “financial services, government, technology, and media,” and its highest-profile work — the December 2018 report for the U.S. Senate Select Committee on Intelligence on Russian 2016-election activity — was commissioned. That is not a benign-by-default purpose; a private firm whose graphs become the public record of who-coordinates-with-whom, paid by the platforms and governments those graphs describe, has a structural conflict built into the business model. Who pays for the map is part of what the map is. The deeds below are weighed against that conflict, not against a neutral-academic prior.

Consequential actions, especially where it cost something. The values-under-cost test for a government contractor is whether it will publish a finding that embarrasses a government customer — and on the one clean documented instance, Graphika did. In August 2022 Graphika co-authored, with the Stanford Internet Observatory, “Unheard Voice: Evaluating five years of pro-Western covert influence operations” (published Aug 24 2022), which documented a covert network advancing U.S. and allied narratives in the Middle East and Central Asia — an operation subsequently linked in reporting to U.S. Special Operations Command, and one that triggered a Pentagon audit of its clandestine information warfare.

That Graphika has itself been described in reporting as a “Pentagon information warfare contractor” is what makes the action load-bearing: a firm that sells to the government nonetheless published a map embarrassing to a government client. That is the conflict being tested and, on this instance, not failing. But one counter-example does not dissolve a structural conflict — the firm’s product is still shaped by who commissions it, and most of its commercial work is not published for outside scrutiny at all.

Leadership choices. Graphika is a revolving-door hinge, and its leadership ledger is the exhibit. Founder John Kelly remains Executive Chairman. Camille François served as Chief Innovation Officer and authored the field-standard ABC (Actor–Behavior–Content) framework before moving to a frontier lab and then the presidency of ROOST; Ben Nimmo carried cross-platform investigations from DFRLab → Graphika → Meta → OpenAI.

The placement direction is the opposite of SIO’s: where the contested research lab lost its people out of the apparatus, Graphika placed its people into it — the framework author into a frontier lab and a safety org, the lead investigator into the two largest platforms running the systems the next operations will target. A commercial map-maker whose graduates become the labs’ threat-intelligence leads holds a quieter, more durable kind of reach than any single report.

CONDUCT verdict: CONFLICTED — a commercial influence-mapping firm with a conflict built into its model (it sells the authoritative map to the governments and platforms it maps), partly mitigated by a documented willingness to publish against a government client (“Unheard Voice”), but never resolved, and amplified by a revolving door that placed its framework author and lead investigator inside the labs and platforms its maps describe.

Reach Assessment

Institutional: Graphika’s maps became the picture the platforms, the press, and a Senate committee reasoned inside — reach measured in adopted graphs, not in any single takedown. Memetic: Through François’s ABC framework, the firm’s grammar — Actor, Behavior, Content — propagated into how the entire field decomposes a disinformation campaign; owning the grammar is upstream of every sentence spoken in it. Civilizational: Graphika does not build AI systems or write their rules. It built the commercial instrument that decides what the public picture of online coordination looks like — and its people carried that instrument into the labs whose systems the next operations will run on. The breach reach of an influence map is wide because a graph travels as fact: it is screenshotted, cited, and entered into the record long after anyone checks who drew the boundary.


Sources: Our Story — Graphika; John W. Kelly — Graphika; Ex-Google exec Camille François investigated Russian trolls — CNBC, Jan 27 2019; Graphika Chief Innovation Officer Camille François Named MIT Technology Review 2019 “Innovator Under 35” — PR Newswire; Unheard Voice: Evaluating five years of pro-Western covert influence operations — Graphika & Stanford Internet Observatory, Aug 24 2022; Fewer Bots, More Ads: The Pentagon’s Evolving Online Influence Campaigns — Lawfare.

ATK 7 ACCELERATION
DEF 7 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.