Screener/Hugging Face

Hugging Face

Leading AI model hosting platform with strong community engagement but complex dual-use hosting challenges.

HQ🇺🇸 US
Est2016
Size201-500
EU AI ActGPAI
huggingface.co
Score
45.0 / 100
Evidence
8 items
Confidence
medium

Developing safety practices - core foundations in place with room for improvement.

Strengths:External Engagement
Weaknesses:Governance Maturity, Technical Safety, Risk Assessment, Regulatory Readiness
Focus Areas
AI platformmodel hubopen source AITransformers

Strengths

  • High evidence coverage (8 items)

Risks

  • Risk requires attention
Table of Contents

Security Assessment

Security-relevant indicators for vendor evaluation

Security Posture
43
TS-01dim: 48
Red Teaming & Pre-deployment Testing
Adversarial testing before deployment
TS-05dim: 48
Robustness & Adversarial Resilience
Resistance to adversarial attacks
RA-01dim: 38
Sector-Specific Risk Assessment
Risk analysis for deployment context
RA-03dim: 38
Dual-Use & Misuse Risk
Dangerous capability awareness
RA-07dim: 38
Incident History & Track Record
Past incidents and response quality
EE-04dim: 55
Vulnerability Disclosure Program
Bug bounty or CVE reporting process
Incident History
Hugging Face incident records sourced from AIAAIC Repository and public reporting.
Integration: AIAAIC, OECD AI Incidents Monitor
Third-Party Audits
External audit reports, SOC 2 attestations, and ISO certifications verified where published.
Sources: Company filings, registry lookups
CVE & Disclosures
Known vulnerabilities and security advisories from NVD, GitHub Security Advisories, and vendor pages.
Sources: NVD, GHSA, vendor disclosure pages

Dimension Breakdown

GM
Governance Maturitypreliminary
Published policies, corporate structure, safety mandate, whistleblowing, executive commitment.
42
TS
Technical Safetypreliminary
Benchmarks, adversarial robustness, fine-tuning safety, watermarking, model cards, research output.
48
RA
Risk Assessmentpreliminary
Dangerous capability evaluations, thresholds, external testing, bug bounty, halt conditions.
38
RR
Regulatory Readinesspreliminary
ISO 42001, EU AI Act compliance, GPAI obligations, international commitments, incident reporting.
42
EE
External Engagementpreliminary
Survey participation, research support, transparency, behavior specs, open-source contributions.
55

Social Impact & Safety Profile

Moderate

Hugging Face is the largest open AI model and dataset hub, hosting hundreds of thousands of models. Strong community engagement and transparency culture. Publishes model cards, dataset cards, and safety documentation. However, hosting third-party models creates content moderation and dual-use challenges at scale. Has implemented model gating and content policies but enforcement at scale remains difficult.

model hosting safetycommunity governanceopen source AIcontent moderation

Peer Comparison

AI Underwriting Company
A-70

Governance Tooling

Compare
Gray Swan
B+64.7

Evaluations & Benchmarking

Compare
Haize Labs
B56.7

Evaluations & Benchmarking

Compare
Goodfire
B56.3

Interpretability

Compare
Compare side-by-side →

Data Sources & Methodology

Scoring methodology v0.1 · 40 indicators · 6 frameworks

Last assessment: 2026-03-23 · Confidence: medium · Evidence: 8 items

NIST AI RMF · EU AI Act · ISO 42001 · FLI AI Safety Index · MLCommons AILuminate · METR

How is this company scored? →Full methodologyView in Screener

Scores reflect publicly available information. A low score may indicate limited transparency rather than poor safety practices.