evilrobots.lol / tech / evaluator record

Gryphon Scientific

What it was engaged to test, on what terms, what it found in the document's own words, and what happened afterwards. Every line is a record; every record links to the document it was read from.

▣ THE RECORD for-profit Bio contractors STORE 2026-10-23
  • Foundednot stated
  • HQMaryland
  • Registry idnone cited (company)
  • TestsBiological-risk red teaming and uplift-study design (tasks, rubrics, graders) for Anthropic and OpenAI; UK AI Security Institute supplier.
  • Money, as the roster puts itPrivate firm; UK DSIT award £392,486.78 after variation (sole tenderer, 2024) and an NIH contract; assets bought by Deloitte Consulting (April 2024).
  • Filesdossier · observatory, sliced to it · graph
Labs tested
OpenAI Anthropic
Models
GPT-5.5GPT-5.6GPT-6 Astra
Span of engagements
26 July 2023 – August 2026
Terms and findings
1 of 4 documents state a term · 1 quote a finding
Money in, US dollars, stated single-recipient rows
no stated figure 2 in · 0 outfrom UK DSIT
Seats, by side
no seat names it
Afterwards
no later event on record

Each count is the number of store records that name this evaluator. An engagement is one lab document naming it for one model; an outcome is an event on the record afterwards; the money, seats and moves are the observatory's rows filtered to it. A cell that cannot be filled says not stated, and each table ends with how many of its rows state each field.

▣ SLICE 6 / 6 records
Lab
Model
Engagement
Terms
Funder
year, month or day

Filters combine: any of the values ticked within a dimension, all of the dimensions at once. A table that has no column for a dimension you have set says so instead of showing rows that cannot match. The address bar carries the slice.

What it was engaged to test 4 engagements, OpenAI, Anthropic

In date order, each from the lab document that names the evaluator. Under each: the four terms that document states (access, model stage, publication, payment) and the finding in its words. A term the document is silent on reads not stated; the tags beside the terms are read off their words and are what the console filters on.

  1. 26 July 2023 Anthropic · no model named pre-deployment testing
    Document: Anthropic Tested Claude for biological risk (2023)
    • Access window"Over the past six months, we spent more than 150 hours with top biosecurity experts red teaming and evaluating our model's ability to output harmful biological information"
    • Model stage"The experts used a bespoke, secure interface to our model without the trust and safety monitoring and enforcement tools that are active on our public deployments"
    • Publicationnot stated
    • Paymentnot stated
    Finding, in the document's words"working with experts, we found that models might soon present risks to national security, if unmitigated"
    • Access windownot stated
    • Model stagenot stated
    • Publicationnot stated
    • Paymentnot stated

    Finding: not stated in the document

    • Access windownot stated
    • Model stagenot stated
    • Publicationnot stated
    • Paymentnot stated

    Finding: not stated in the document

    • Access windownot stated
    • Model stagenot stated
    • Publicationnot stated
    • Paymentnot stated

    Finding: not stated in the document

Dated 4 of 4Model named 3 of 4Terms stated 1 of 4Finding quoted 1 of 4

The terms, as the documents state them

The same engagements as a grid, one row per document and model, so the silences line up: a document that gives an access window but no payment term, or a finding but no publication term, reads that way here.

DocumentModelAccess windowModel stagePublicationPaymentFindingRecord
AnthropicAnthropic · 26 July 2023no model named"Over the past six months, we spent more than 150 hours with top biosecurity experts red teaming and evaluating our model's ability to output harmful biological information""The experts used a bespoke, secure interface to our model without the trust and safety monitoring and enforcement tools that are active on our public deployments"not statednot statedquotedff0abf8835
GPT-5.5 system cardOpenAI · 23 April 2026GPT-5.5not statednot statednot statednot statednot stated5571273e1d
GPT-5.6 system cardOpenAI · 9 July 2026GPT-5.6not statednot statednot statednot statednot stated9f8853143e
GPT-6 Astra system cardOpenAI · August 2026GPT-6 Astranot statednot statednot statednot statednot statedcfa2c98335

Access window 1 of 4Model stage 1 of 4Publication 0 of 4Payment 0 of 4Finding, quoted 1 of 4

What happened afterwards 0 on record

Events a party published after an engagement: an incident on a model this evaluator's engagement covered, a correction or added disclosure, a later finding that cut against an earlier one, a lab's statement that it acted or did not act on the finding, or a move from this evaluator to a lab it graded. A row the source attaches to this evaluator by name is marked named; one that reaches it through the model it tested is marked via the model. The kind is what the record states, not why.

KindDeveloperModelWhat the record statesWhenSourceId
no later event on record

Dated 0 of 0Model named 0 of 0Tied to an engagement 0 of 0Evaluator named by the record 0 of 0

The money on record 2 rows

Rows of the observatory's money table that name this evaluator as recipient (in) or as funder (out), each as the record states it: a grant table line, a contract notice, a Form 990 schedule or the party's own disclosure. A figure quoted in prose is marked and never summed.

DirectionFunderRecipientAmount, as recordedWhenRecordSourceId
inUK DSITGryphon Scientific£267,618£267,618 (varied to £392,486.78, March 2024)March 2024contract noticeFind a Tender; Find a Tender; Contracts Finder; Contracts Finder; Contracts Finder; Find a Tender, variationbb860f8d21
inUK DSITGryphon Scientific£267,618£267,618, varied to £392,486.78 (March 2024)Analysis and research for the institute's AI safety programmesMarch 2024contract noticeFind a Tender; Find a Tender, variation1244809986

Amount stated 2 of 2Dated 2 of 2Purpose stated 1 of 2Funder profiled 0 of 2Recipient on the roster 2 of 2

Seats and moves 0 seats · 0 moves

Seats read off a team page, a governance page, an appointment notice or a filing that name this organization, and the dated moves into or out of it, from the observatory's own tables.

PersonSideSeat, as the record states itRoleDatesSourceId
no seat on record names this evaluator

Role read 0 of 0Organization read 0 of 0Organization profiled 0 of 0Dated 0 of 0

Moves

PersonFromToWhenSourceId
no move on record names this evaluator

From profiled 0 of 0To profiled 0 of 0Person profiled 0 of 0Dated 0 of 0