AI21 Labs · AI21 Labs Privacy Policy

AI Model Training Use of User Inputs

High severity
Share 𝕏 Share in Share 🔒 PDF

What it is

AI21 may use the text you type into their AI tools — your prompts and the AI's responses — to improve and train their AI models. They say they'll try to reduce personally identifying information, but there is no guarantee your content won't be used.

Consumer impact (what this means for users)

Text inputs and outputs you submit through AI21's platform or API may be used to train future AI models, meaning confidential business information, personal queries, or proprietary content could be incorporated into AI21's systems.

What you can do

⚠️ These actions may provide transparency or partial mitigation but may not fully address the underlying issue. Effectiveness varies by jurisdiction and individual circumstances.
  • Delete Your Data
    Within 30 days
    Email privacy@ai21.com requesting that your data be excluded from AI model training or deleted entirely. Include your account email and specify the nature of your request. AI21 is required to respond within applicable statutory timeframes (30 days under GDPR, 45 days under CCPA).

Cross-platform context

See how other platforms handle AI Model Training Use of User Inputs and similar clauses.

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

Why it matters (compliance & risk perspective)

Your private queries and creative or business inputs submitted to AI21's platform may become training data for future AI systems, raising concerns about confidentiality, intellectual property, and data sovereignty.

View original clause language
We may use the information you provide to us, including information you submit via our Services (such as prompts, inputs, and outputs), to train, improve, and develop our AI models and Services. We will use reasonable measures to minimize the use of information that identifies you personally when using data for these purposes.

Institutional analysis (Compliance & legal intelligence)

(1) REGULATORY FRAMEWORK: This provision implicates GDPR Art. 6(1) (lawful basis for processing — legitimate interests or consent required for training data use), GDPR Art. 5(1)(b) (purpose limitation — secondary use for training must be compatible with original collection purpose), CCPA/CPRA §1798.120 (right to opt out of sale/sharing), and GDPR Art. 22 where outputs influence automated decisions. The EU AI Act imposes transparency obligations on general-purpose AI model providers regarding training data under Art. 53. Enforcement authorities include EU national data protection authorities, California Privacy Protection Agency, and FTC (Section 5 for deceptive practices). (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 has authority over unfair or deceptive data practices under Section 5, including undisclosed use of consumer data for AI training purposes.
    File a complaint →

Provision details

Document information
Document
AI21 Labs Privacy Policy
Entity
AI21 Labs
Document last updated
April 29, 2026
Tracking information
First tracked
April 30, 2026
Last verified
April 30, 2026
Record ID
CA-P-004111
Document ID
CA-D-00460
Evidence Provenance
Source URL
Wayback Machine
SHA-256
4abc7ff0d7779bee955894a99670d17aadf5332ce2786437f3a3b85a2497adc3
Verified
✓ Snapshot stored   ✓ Change verified
How to Cite
ConductAtlas Policy Archive
Entity: AI21 Labs | Document: AI21 Labs Privacy Policy | Record: CA-P-004111
Captured: 2026-04-30 06:15:21 UTC | SHA-256: 4abc7ff0d7779bee…
URL: https://conductatlas.com/platform/ai21-labs/ai21-labs-privacy-policy/ai-model-training-use-of-user-inputs/
Accessed: May 2, 2026
Classification
Severity
High
Categories

Other provisions in this document