AI Training Data License Agreement

AI TRAINING DATA LICENSE AGREEMENT

TABLE OF CONTENTS

1. Parties
2. Recitals
3. Definitions
4. License Grant
5. Permitted and Prohibited Uses
6. Data Quality and Documentation
7. Intellectual Property Rights
8. Copyright and Legal Compliance
9. Privacy and Data Protection
10. Compensation
11. Representations and Warranties
12. Indemnification
13. Limitation of Liability
14. Term and Termination
15. Post-Termination Obligations
16. General Provisions
17. Signatures

1. PARTIES

This AI Training Data License Agreement ("Agreement") is entered into as of [__/__/____] ("Effective Date") by and between:

Licensor: [________________________________] ("Licensor")
Address: [________________________________]

Licensee: [________________________________] ("Licensee")
Address: [________________________________]

2. RECITALS

WHEREAS, Licensor owns or controls certain data assets described herein; and

WHEREAS, Licensee desires to obtain a license to use such data for the purpose of training, developing, validating, and/or fine-tuning artificial intelligence and machine learning models;

NOW, THEREFORE, the parties agree as follows:

3. DEFINITIONS

"Licensed Data" means the datasets described in Exhibit A, including all content, annotations, labels, metadata, and associated documentation.

"AI Model" means any artificial intelligence or machine learning model, algorithm, neural network, or system trained, developed, fine-tuned, or validated using the Licensed Data.

"Derivative Work" means any modification, compilation, transformation, translation, or adaptation of the Licensed Data, including embeddings, tokenizations, vector representations, and statistical models derived from the Licensed Data.

"Output" means any content, prediction, classification, recommendation, or other result generated by an AI Model.

"Training" means any process by which an AI Model learns patterns, representations, or parameters from data, including pre-training, fine-tuning, reinforcement learning, and retrieval-augmented generation indexing.

4. LICENSE GRANT

4.1 Subject to the terms of this Agreement, Licensor grants Licensee a [select one]:

☐ Non-exclusive, non-transferable, non-sublicensable license
☐ Exclusive license within the Field of Use defined in Section 5
☐ Non-exclusive license with sublicense rights subject to Licensor's prior written consent

to use the Licensed Data solely for Training AI Models as specified in Section 5.

4.2 Territorial Scope: [Select one:]
☐ Worldwide
☐ Limited to: [________________________________]

4.3 Duration of License: The license granted herein is effective for the Term set forth in Section 14.

5. PERMITTED AND PROHIBITED USES

5.1 Permitted Uses:

☐ Training AI Models within the following Field of Use: [________________________________]
☐ Internal research and development
☐ Validation and benchmarking of AI Model performance
☐ Fine-tuning pre-trained models for Licensee's commercial products
☐ Other: [________________________________]

5.2 Prohibited Uses:

☐ Redistribution, resale, or sublicensing of the Licensed Data in raw or minimally transformed form
☐ Use of Licensed Data to train AI Models that generate content competing directly with Licensor's products
☐ Reverse engineering, de-anonymization, or re-identification of individuals in the Licensed Data
☐ Use in any manner that violates applicable law, including copyright law (17 U.S.C. § 101 et seq.), the EU Copyright Directive (Directive (EU) 2019/790, Art. 4), or privacy regulations
☐ Combining Licensed Data with data obtained through web scraping in violation of terms of service or applicable law
☐ Use for development of weapons systems, mass surveillance, or social scoring

6. DATA QUALITY AND DOCUMENTATION

6.1 Licensor shall provide a data card or dataset documentation sheet as Exhibit B, describing:

• Data sources, collection methodology, and temporal scope
• Annotation schema and labeling methodology
• Known limitations, biases, or quality issues
• Demographic composition (where applicable)
• Data format and technical specifications

6.2 Licensor shall promptly notify Licensee of any material errors, contamination, or quality issues discovered in the Licensed Data after delivery.

7. INTELLECTUAL PROPERTY RIGHTS

7.1 Licensor retains all right, title, and interest in and to the Licensed Data, including all copyrights, database rights (Directive 96/9/EC), and other intellectual property rights.

7.2 Licensee retains all right, title, and interest in AI Models trained using the Licensed Data, provided that such AI Models do not incorporate the Licensed Data in a form that is extractable or reconstructable.

7.3 Ownership of Derivative Works: [Select one:]
☐ Licensor retains ownership of all Derivative Works
☐ Licensee owns Derivative Works, subject to Licensor's rights in the underlying Licensed Data
☐ Joint ownership as specified in Exhibit C

7.4 Outputs generated by AI Models trained on the Licensed Data are not subject to Licensor's copyright claims, provided the Outputs do not constitute substantially similar reproductions of the Licensed Data.

8. COPYRIGHT AND LEGAL COMPLIANCE

8.1 Licensor represents that it holds all necessary rights to license the Licensed Data for AI Training purposes, including any required permissions from individual copyright holders.

8.2 Where the Licensed Data includes works subject to the EU Copyright Directive, Art. 4 text and data mining exception, Licensor confirms that rights holders have not exercised their opt-out reservation under Art. 4(3).

8.3 Licensee shall implement reasonable technical measures to prevent the AI Model from memorizing and reproducing substantial portions of the Licensed Data in its Outputs.

8.4 Licensee shall maintain records of the Licensed Data used in Training, sufficient to comply with EU AI Act transparency requirements (Reg. (EU) 2024/1689, Art. 53(1)(d)) and any applicable regulatory inquiry.

9. PRIVACY AND DATA PROTECTION

9.1 If the Licensed Data contains personal data as defined by GDPR Art. 4(1) or CCPA Cal. Civ. Code § 1798.140(v), the parties shall execute a Data Processing Addendum as Exhibit D.

9.2 Licensor represents that all personal data in the Licensed Data was collected with valid legal basis under applicable privacy law and may be lawfully used for AI Training.

9.3 Licensee shall not attempt to re-identify anonymized or pseudonymized individuals in the Licensed Data.

10. COMPENSATION

10.1 License Fee: Licensee shall pay Licensor:

☐ One-time license fee of $[________________________________]
☐ Annual license fee of $[________________________________]
☐ Per-record / per-unit fee of $[________________________________] per [________________________________]
☐ Revenue share of [____]% of net revenue derived from AI Models trained on the Licensed Data
☐ Other: [________________________________]

10.2 Payment terms: Net [____] days from invoice date.

11. REPRESENTATIONS AND WARRANTIES

11.1 Licensor Warranties:
(a) Licensor has the legal right and authority to grant the license herein;
(b) The Licensed Data does not, to Licensor's knowledge, infringe any third-party intellectual property rights;
(c) The Licensed Data conforms materially to the description in Exhibit A and the data card in Exhibit B.

11.2 Licensee Warranties:
(a) Licensee will use the Licensed Data solely within the scope of the license granted;
(b) Licensee will comply with all applicable laws in its use of the Licensed Data.

12. INDEMNIFICATION

12.1 Licensor shall indemnify Licensee against third-party claims alleging that the Licensed Data infringes intellectual property rights or was collected in violation of privacy law.

12.2 Licensee shall indemnify Licensor against claims arising from Licensee's use of the Licensed Data outside the scope of this Agreement or Licensee's AI Model Outputs.

13. LIMITATION OF LIABILITY

13.1 EXCEPT FOR INDEMNIFICATION OBLIGATIONS AND BREACHES OF IP RESTRICTIONS, NEITHER PARTY'S AGGREGATE LIABILITY SHALL EXCEED THE TOTAL LICENSE FEES PAID OR PAYABLE UNDER THIS AGREEMENT.

13.2 NEITHER PARTY SHALL BE LIABLE FOR INDIRECT, INCIDENTAL, SPECIAL, CONSEQUENTIAL, OR PUNITIVE DAMAGES.

14. TERM AND TERMINATION

14.1 Term. This Agreement commences on the Effective Date and continues for [____] year(s), unless earlier terminated.

14.2 Termination for Breach. Either party may terminate upon [____] days' written notice of material breach if the breach remains uncured.

14.3 Termination for Insolvency. Either party may terminate immediately if the other party becomes insolvent or files for bankruptcy.

15. POST-TERMINATION OBLIGATIONS

15.1 Upon termination, Licensee shall cease all new Training using the Licensed Data and shall delete or return all copies of the Licensed Data within [____] days.

15.2 [Select one:]
☐ Licensee may continue to use AI Models already trained on the Licensed Data prior to termination ("Survival of Trained Models")
☐ Licensee must cease all use of AI Models trained on the Licensed Data upon termination

16. GENERAL PROVISIONS

16.1 Governing Law. This Agreement shall be governed by the laws of the State of [________________________________].

16.2 Dispute Resolution. [Select one:]
☐ Exclusive jurisdiction in courts of [________________________________]
☐ Binding arbitration under [________________________________] rules

16.3 Confidentiality. Each party shall maintain the confidentiality of the other party's proprietary information for [____] years following disclosure.

16.4 Entire Agreement. This Agreement constitutes the entire agreement between the parties regarding its subject matter.

17. SIGNATURES

LICENSOR:

Signature: [________________________________]
Printed Name: [________________________________]
Title: [________________________________]
Date: [__/__/____]

LICENSEE:

Signature: [________________________________]
Printed Name: [________________________________]
Title: [________________________________]
Date: [__/__/____]

EXHIBITS

• Exhibit A: Licensed Data Description
• Exhibit B: Data Card / Dataset Documentation
• Exhibit C: Derivative Works Ownership (if applicable)
• Exhibit D: Data Processing Addendum (if personal data)

This agreement template does not constitute legal advice. Parties must consult qualified legal counsel before execution.