Screener/Google DeepMind

Google DeepMind

Leading AI research lab with extensive safety research but governance challenges from rapid commercialization.

HQ🇬🇧 GB
Est2010
Size1000+
EU AI ActGPAI Systemic
deepmind.google
Score
65.0 / 100
Evidence
11 items
Confidence
high

Strong safety posture with established governance frameworks and active risk management.

Strengths:Governance Maturity, Technical Safety, Risk Assessment, Regulatory Readiness, External Engagement
Focus Areas
frontier modelsgeminialignmentsafety researchresponsible ai

Strengths

  • Technical score (72) indicates strong practices
  • High evidence coverage (11 items)

Risks

No significant risks identified

Table of Contents

Security Assessment

Security-relevant indicators for vendor evaluation

Security Posture
65
TS-01dim: 72
Red Teaming & Pre-deployment Testing
Adversarial testing before deployment
TS-05dim: 72
Robustness & Adversarial Resilience
Resistance to adversarial attacks
RA-01dim: 58
Sector-Specific Risk Assessment
Risk analysis for deployment context
RA-03dim: 58
Dual-Use & Misuse Risk
Dangerous capability awareness
RA-07dim: 58
Incident History & Track Record
Past incidents and response quality
EE-04dim: 67
Vulnerability Disclosure Program
Bug bounty or CVE reporting process
Incident History
Google DeepMind 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.
68
TS
Technical Safetypreliminary
Benchmarks, adversarial robustness, fine-tuning safety, watermarking, model cards, research output.
72
RA
Risk Assessmentpreliminary
Dangerous capability evaluations, thresholds, external testing, bug bounty, halt conditions.
58
RR
Regulatory Readinesspreliminary
ISO 42001, EU AI Act compliance, GPAI obligations, international commitments, incident reporting.
60
EE
External Engagementpreliminary
Survey participation, research support, transparency, behavior specs, open-source contributions.
67

Social Impact & Safety Profile

Moderate

Google DeepMind maintains one of the largest AI safety research teams globally, publishing extensively on alignment, interpretability, and evaluation. Their Frontier Safety Framework outlines safety protocols for advanced models. However, internal tensions between safety and product timelines have been reported, and the pace of Gemini deployments sometimes outstrips governance processes.

alignment researchfrontier safety frameworkai evaluationresponsible deployment

Peer Comparison

Anthropic
A-72.8

Alignment Research

Compare
OpenAI
C+46.2

Alignment Research

Compare
Meta
C43.8

Alignment Research

Compare
xAI
D25

Foundation Models

Compare
Compare side-by-side →

Data Sources & Methodology

Scoring methodology v0.1 · 40 indicators · 6 frameworks

Last assessment: 2026-03-23 · Confidence: high · Evidence: 11 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.