Fuel50
Fuel50 is an AI internal talent marketplace and workforce-transformation platform that matches employees to internal roles, gigs, mentors, learning and career paths, and fit-scores and ranks internal candidates for open roles and succession pipelines, using a skills ontology curated by in-house I/O psychologists.
HireAIScore rates Fuel50 58 out of 100 (grade F, Substantial gaps) under rubric v1.0, last reviewed September 25, 2026. Fuel50 ranks 7 of 86 vendors rated in Candidate sourcing AI, against a category average of 46. That score is drawn from 28 evidence items across 7 rubric criteria for Fuel50 — 24 with a cited source, 4 recording that nothing was located.
§ 01 - Company facts
- Legal name
- Fuel50 Inc.
- Founded
- 2011
- Headquarters
- United States · US
- Pricing tier
- Enterprise
- Market side
- Employer-side (AEDT)
- Categories
- SourcingAssessment
- Website
- fuel50.com
These are company details, not findings. They are editable by a verified representative and carry no weight in the score, which is built only from the cited evidence below.
§ 02 - Score breakdown
§ Score breakdown
Category scoring
Weighted contribution shown to the right of each bar.
- 01
Article 11 Technical Documentation
Weight 20%70
+14.0 · category avg 48
- 02
Bias Audit Transparency
Weight 18%68
+12.2 · category avg 40
- 03
FRIA Support
Weight 15%32
+4.8 · category avg 31
- 04
Data Governance Disclosure
Weight 15%64
+9.6 · category avg 53
- 05
Human Oversight Design
Weight 12%61
+7.3 · category avg 56
- 06
Post-Market Monitoring
Weight 12%40
+4.8 · category avg 40
- 07
Customer Documentation
Weight 8%62
+5.0 · category avg 57
Category avg is the mean raw score on that criterion across the 86 Candidate sourcing AI vendors in scope of this rubric, this one included.
§ 03 - Strongest · weakest
Strongest category
Article 11 Technical Documentation
Raw score 70 · contributes 14.0 to total.
70 against a 48 category average
Weakest category
FRIA Support
Raw score 32 · contributes 4.8 to total.
32 against a 31 category average
§ 04 - Cited evidence
Download diligence record→§ Evidence
Cited per category
Every score is backed by at least one cited piece of evidence.
Evidence ledger
- Items
- 28
- Documentation
- 18
- Audit report
- 2
- Public statement
- 4
- Absence
- 4
- With a source URL
- 24 of 28
- Source hostnames
- 4
- Fewest items
- 3
- Bias Audit Transparency
These figures measure how thoroughly Fuel50 was reviewed, not how Fuel50 performed — an absence row, recording that nothing was located, is counted like any other item.
Article 11 Technical Documentation
4 items70
- DocumentationCaptured Sep 25, 2026
The Trust Center shows a publicly viewable ISO/IEC 42001:2023 certificate from ANAB-accredited Sensiba LLP to Career Engagement Group Limited t/a Fuel50 (no. FUE-AIMS-20260623, issued 23 June 2026, valid to 22 June 2029), scoped to the design, development, deployment, monitoring and improvement of AI features in its talent-marketplace platform.
- DocumentationCaptured Sep 25, 2026
A publicly viewable 15-page Responsible AI Policy (version 1, approved 16 February 2026) sets principles, risk tiers, prohibited uses and internal AI Impact Assessments, cites NIST AI RMF, ISO/IEC 42001 and the OWASP LLM Top 10, and says every recommendation is traceable to specific algorithmic rules.
- DocumentationCaptured Sep 25, 2026
The trust-center AI FAQ describes the six AI techniques in use (supervised ML for matching trained on expert-created datasets, cross-encoder and re-ranker models, heuristic matching, third-party LLMs with opt-out, a proprietary NLP skills parser, semantic analysis) and says most models are built in-house.
- AbsenceCaptured Sep 25, 2026
No public system or model card, intended-purpose and limitations statement, accuracy or validation metrics, or instructions for use were found on fuel50.com, the Trust Center resources and FAQ, or the May 2026 MSA; the RAI Policy's model-factsheet template is internal and docs.fuel50.com redirects to a login page.
Bias Audit Transparency
3 items68
- Audit reportCaptured Sep 25, 2026
Publicly viewable Holistic AI NYC LL 144 audit of FuelMobility and Succession, which fit-scores and ranks employees for open roles and succession: 755,431 historical records (1 Jan 2020 to 4 Jul 2025), impact ratios 0.94-1.00 standalone and 0.91-1.00 intersectional for reported groups, and no audit completion date shown.
- Public statementCaptured Sep 25, 2026
Fuel50's FAQ says Holistic AI has audited its Mobility and Succession engines under NYC LL 144 since 2023, repeatedly finding no disparate impact by gender or ethnicity, but only one audit (data to July 2025) is publicly linked.
- DocumentationCaptured Sep 25, 2026
The Trust Center lists 'New York City Local Law 144 – AI Bias Audit' as a publicly viewable resource linking to the Holistic AI summary; no earlier or later audit cycle is listed.
FRIA Support
3 items32
- Public statementCaptured Sep 25, 2026
The trust-center FAQ answers 'No' to whether Fuel50's AI systems are high-risk under the EU AI Act, claiming exceptions for Successions AI (a preparatory task supporting human review) and Opportunity Matching (rule-based skill matching), and says it will review this once the Act fully applies.
- Public statementCaptured Sep 25, 2026
The marketing FAQ, by contrast, says the EU AI Act treats employment AI as high-risk and that deploying Fuel50 means inheriting a compliance posture, but it gives no deployer-obligation guidance.
- AbsenceCaptured Sep 25, 2026
No FRIA template, Article 26/27 deployer guidance or EU AI Act instructions were found on fuel50.com (FAQ, product pages, privacy and legal), in the Trust Center resources and FAQ, or in the May 2026 MSA and DPA; the RAI Policy's AI Impact Assessment template is internal, and the help center and docs site require login.
Data Governance Disclosure
5 items64
- Audit reportCaptured Sep 25, 2026
A publicly viewable Sensiba attestation letter confirms a SOC 2 Type 2 report for Fuel50 Inc covering Security, Availability and Confidentiality from 1 July 2025 to 30 June 2026 (opinion dated 15 July 2026); the full report and penetration test are available on request only.
- DocumentationCaptured Sep 25, 2026
The trust-center AI FAQ says named-entity recognition strips names, locations, organizations, dates, contact details, nationality, religion and political views before any external LLM processing, and that clients can opt out of third-party LLMs and of the use of anonymized data for model improvement.
- DocumentationCaptured Sep 25, 2026
The public subprocessor list names nine providers with processing regions, including AWS hosting in Europe, the US and Singapore, OpenAI for generative AI, and LangChain's LangSmith for monitoring Fuel50's agents.
- DocumentationCaptured Sep 25, 2026
The public MSA with DPA (updated 27 May 2026) bars using client content to train public or cross-client AI models without consent, but lets Fuel50 train algorithms on client role-movement data and on aggregated or anonymized client content.
https://fuel50.com/hubfs/Fuel50%20-%20MSA%20-%20SLA%20-%20DPA%20-%202026%20May.pdf
- DocumentationCaptured Sep 25, 2026
The CSA STAR Registry lists Fuel50 at STAR Level 1 with a CAIQ v4.0.2 self-assessment, created or renewed on 22 April 2026.
https://cloudsecurityalliance.org/star/registry/fuel50/services/fuel50-com
Human Oversight Design
4 items61
- DocumentationCaptured Sep 25, 2026
The trust-center FAQ describes human-in-the-loop controls: human approval of all AI-generated content through staged Draft, Review and Approval workflows, audit trails with timestamps and user attribution, required reasons for declined changes, and per-feature AI disablement.
- Public statementCaptured Sep 25, 2026
The privacy policy (updated 11 May 2026) says Fuel50 does not use AI to make decisions with legal or similarly significant effects on individuals without appropriate human involvement.
- DocumentationCaptured Sep 25, 2026
The Succession page describes decision-support outputs for HR and leaders: readiness tiers ('Ready now', 'Ready soon', 'Emergency coverage'), documented skills-matching logic, and shared pipeline ownership with notes and annotations.
- DocumentationCaptured Sep 25, 2026
The Responsible AI Policy prohibits unfair and unexplainable automated promotion or termination decisions and says every recommendation is traceable to specific algorithmic rules.
Post-Market Monitoring
5 items40
- AbsenceCaptured Sep 25, 2026
No public status page or incident history was found: status.fuel50.com does not resolve; fuel50.statuspage.io, fuel50.instatus.com and fuel50.betteruptime.com redirect to the providers' homepages; fuel50.freshstatus.io returns a 404 page; and fuel50.com serves no security.txt.
- DocumentationCaptured Sep 25, 2026
The public MSA commits to 99.5% monthly uptime, at least two business days' notice of maintenance, and notifying the client of a security incident within 48 hours.
https://fuel50.com/hubfs/Fuel50%20-%20MSA%20-%20SLA%20-%20DPA%20-%202026%20May.pdf
- DocumentationCaptured Sep 25, 2026
The Responsible AI Policy defines AI incidents, routes them through the incident-management process and says materialized impacts will be recorded internally and publicly reported, but no public AI-incident log or model changelog was found.
- DocumentationCaptured Sep 25, 2026
The subprocessor list discloses that LangChain's LangSmith observability service monitors and evaluates the behaviour and performance of Fuel50's proprietary agents.
- DocumentationCaptured Sep 25, 2026
The contact page lists a 24/7 help line and help@fuel50.com; privacy@fuel50.com is the only compliance contact on the Trust Center, and no dedicated security or vulnerability-reporting address was found.
Customer Documentation
4 items62
- DocumentationCaptured Sep 25, 2026
The buyer FAQ covers AI bias testing, the Holistic AI audits, EU AI Act and NYC LL 144 positioning, SOC 2, data residency, the DPA and the Trust Center.
- DocumentationCaptured Sep 25, 2026
The Privacy & Legal page publishes the privacy policy (updated 11 May 2026, with an AI section and DPF commitments) and links the full MSA, SLA and DPA as a public PDF.
- DocumentationCaptured Sep 25, 2026
The Mobility page explains AI matching of employees to open roles and career paths on skills, readiness and aspirations, with role-based permissions for HR, talent acquisition and hiring managers.
- AbsenceCaptured Sep 25, 2026
No customer-facing guidance on employer duties (NYC LL 144 notices, Illinois HB 3773 notice, EU AI Act deployer obligations), instructions for use or limitations statement was found on fuel50.com or the Trust Center; fuel50.freshdesk.com and docs.fuel50.com redirect to login pages.
§ 05 - Editorial notes
Company overview
Fuel50 is an AI internal talent marketplace and workforce-transformation platform. Its Mobility, Succession and Talent Marketplace modules match employees to internal roles, gigs, mentors, learning and career paths, and rank internal candidates for open roles and succession pipelines, using an ontology of more than 5,000 skills curated by in-house I/O psychologists. It was founded in Auckland, New Zealand by Anne Fulton (CEO) and Jo Mills (President). It gives 2011 as its founding year, its New Zealand operating company, Career Engagement Group Limited, dates from 2003, and its corporate base is now Laguna Niguel, California. On 16 September 2026 Phoenix Education Partners (NYSE: PXED) agreed to buy Fuel50 Inc. for about $31.5 million in cash plus up to $8.5 million in earnouts. The deal was expected to close within 30 days, after which Fuel50 will run as a separate brand with Jo Mills as CEO. Pricing is unpublished and set per employee and per module, and named customers such as KeyBank and AIB Group are large enterprises.
Regulatory exposure
Fuel50's exposure sits on the promotion side of employment, which every regime here covers. Its FuelMobility and Succession engine fit-scores and ranks employees for open roles and succession. Under NYC Local Law 144 that is AEDT use for promotion, so NYC employers need an independent bias audit from the past year plus advance notice. Fuel50's public Holistic AI audit helps, but it shows no audit date and its data ends 4 July 2025. Under the EU AI Act, AI used for promotion and work allocation is high-risk (Annex III point 4), with obligations now due 2 December 2027. Fuel50's trust-center FAQ nonetheless says none of its systems is high-risk, citing Article 6(3) exceptions. That is hard to square with Article 6(3)'s rule that Annex III systems which profile people are always high-risk, and with Fuel50's own marketing FAQ, which calls employment AI high-risk. Illinois HB 3773 (from 1 January 2026) covers promotion and selection for training and requires employee notice. Colorado's SB 26-189 disclosure regime starts 1 January 2027. California's FEHA automated-decision rules (since 1 October 2025) and CPPA ADMT rules (from 1 January 2027) also reach promotion decisions.
Path to a higher score
The fastest gains come from publishing what Fuel50 says it already has. Adding the audit date to the Holistic AI summary, posting the earlier audits it says date back to 2023, and releasing a 2026 re-audit would show consecutive public audits and lift bias transparency toward 75 or more. Extending audits to the Career Advisor Agent and gig matching would widen coverage. Its ISO 42001 management system should also yield public system cards for each engine, covering intended purpose, inputs such as peer and leader ratings, limitations and validation metrics, plus instructions for use. Fuel50 should also settle its EU AI Act position: either treat Mobility and Succession as high-risk or publish a defensible Article 6(4) assessment, and give deployers Article 26 guidance, FRIA-ready material and NYC and Illinois notice templates. A public status page and a changelog of model updates and AI incidents would give post-market monitoring its first real public evidence.
§ Regulatory frame
What applies to candidate sourcing ai.
Ranking outputs are inside Annex III §4 when they influence hiring decisions. GDPR Article 22 (automated decision-making) and the GDPR rights around profiling apply independently for EU candidates.
§ Compare
Build any comparison→Fuel50 against its nearest-scoring peers.
In Candidate sourcing AI.
§ Others rated in Candidate sourcing AI
All sourcing vendors→Ranked 7 of 86 by weighted total under rubric v1.0. The ordering is arithmetic on the rubric and carries no view on which tool suits a given hiring process.
- 04Beamery61D
- 05Sense58F
- 06Tezi58F
- 07Fuel50This profile58F
- 08Findem58F
- 09Pinpoint57F
Conflicts of interest
No vendor pays for placement, scoring, or removal. Casework - the consulting firm that operates this directory - provides paid services to some vendors. Any active or recent (within 24 months) commercial relationship is disclosed on the affected vendor profile and the review is reassigned to an independent reviewer. See the full policy on About.
Casework has no commercial relationship with this vendor.