OpenAI · GPT-4o System Card (PDF)

Non-Consensual Intimate Imagery (NCII) Risk

High severity
Share 𝕏 Share in Share 🔒 PDF

What it is

GPT-4o has built-in blocks to prevent creating fake intimate images of real people without their consent, but OpenAI admits these protections are imperfect and the risk remains.

Consumer impact (what this means for users)

Your image or likeness could potentially be used to generate non-consensual intimate imagery using GPT-4o's vision and image capabilities — OpenAI has partial protections in place but explicitly acknowledges they are not fully effective.

Cross-platform context

See how other platforms handle Non-Consensual Intimate Imagery (NCII) Risk and similar clauses.

Compare across platforms →
Need full compliance memos? See Professional →

Why it matters (compliance & risk perspective)

NCII — sometimes called deepfake pornography — causes severe psychological harm to victims, predominantly women, and is now illegal in many jurisdictions; OpenAI's acknowledgment that its mitigations are 'imperfect' is a significant safety and legal exposure disclosure.

View original clause language
We have safeguards to prevent the generation of non-consensual intimate imagery (NCII)... Red teamers noted that GPT-4o's image generation capabilities could be used to generate NCII... We have classifiers and other technical mitigations in place, though we acknowledge that these are imperfect.

Institutional analysis (Compliance & legal intelligence)

(1) REGULATORY FRAMEWORK: This provision implicates the DEEPFAKES Accountability Act (proposed), state NCII statutes (California AB 602/AB 3286, Texas Penal Code §21.165, Virginia Code §18.2-386.2, and over 45 other states with NCII laws), EU AI Act Article 5 on prohibited practices causing harm, GDPR Article 9 on special category data (biometric data used to generate intimate images), and the UK Online Safety Act 2023 (Part 5, non-consensual intimate images). Enforcement authorities include state law enforcement, the UK Ofcom, and national DPAs under GDPR. (2)

🔒

Compliance intelligence locked

Regulatory citations, enforcement risk, and due diligence action items.

Watcher $9.99/mo Professional $149/mo

Watcher: regulatory citations. Professional: full compliance memo.

Applicable agencies

  • FTC
    FTC consumer protection authority covers unfair practices including AI-facilitated non-consensual intimate imagery generation where platform safeguards are acknowledged as imperfect.
    File a complaint →
  • State AG
    State Attorneys General enforce NCII statutes in 45+ states, creating direct enforcement exposure for platforms that fail to adequately prevent AI-generated non-consensual intimate imagery.
    File a complaint →

Provision details

Document information
Document
GPT-4o System Card (PDF)
Entity
OpenAI
Document last updated
March 5, 2026
Tracking information
First tracked
March 10, 2026
Last verified
April 27, 2026
Record ID
CA-P-003148
Document ID
CA-D-00008
Evidence Provenance
Source URL
Wayback Machine
SHA-256
7c23ef53467eea199596abe78511d57ffee1e94b50ef10ac0f7d81df278b5059
Verified
✓ Snapshot stored   ✓ Change verified
How to Cite
ConductAtlas Policy Archive
Entity: OpenAI | Document: GPT-4o System Card (PDF) | Record: CA-P-003148
Captured: 2026-03-10 03:40:55 UTC | SHA-256: 7c23ef53467eea19…
URL: https://conductatlas.com/platform/openai/gpt-4o-system-card-pdf/non-consensual-intimate-imagery-ncii-risk/
Accessed: May 2, 2026
Classification
Severity
High
Categories

Other provisions in this document