IP Due Diligence in AI M&A Transactions – Overview

IP due diligence is the process of reviewing, analyzing, and validating the intellectual property assets of a target company during mergers and acquisitions (M&A). For AI companies, IP due diligence is critical because the value of the business often resides in its intellectual property, including:

AI models and algorithms

Training datasets

Software code and APIs

Patents, trademarks, and copyrights

Trade secrets and proprietary methods

Key objectives of IP due diligence in AI M&A:

Verify ownership of patents, copyrights, and trade secrets

Identify potential infringement claims or licensing issues

Assess IP coverage in products, datasets, and models

Evaluate regulatory and compliance risks (e.g., GDPR, data licensing)

Ensure enforceability of confidentiality agreements and NDAs

Key Steps in AI IP Due Diligence

Patent Review: Examine AI patents (algorithms, models, system implementations) for validity and enforceability.

Copyright Review: Ensure that code, datasets, and APIs used are licensed or original.

Trade Secret Assessment: Check NDAs and internal controls to protect proprietary AI methods.

Licensing & Contracts: Identify open-source, third-party libraries, and software licenses and their obligations.

Litigation Risk: Investigate ongoing or potential IP disputes.

Key Cases Related to IP Due Diligence in AI / Technology M&A

While AI-specific M&A cases are limited due to confidentiality, many tech and software cases provide insights:

1. Waymo v. Uber (2017–2018)

Facts: During M&A and hiring transitions, ex-Waymo employees joined Uber, allegedly taking trade secrets in autonomous vehicle AI.

Issue: Due diligence on IP ownership and employee obligations failed.

Holding: Uber settled for $245 million in stock and agreed to safeguards.

Significance: Highlights the importance of scrutinizing employee movement, NDAs, and trade secrets in AI M&A.

*2. Google v. Levandowski / Otto Trucking (2017)

Facts: Levandowski left Google self-driving division to form Otto; he allegedly downloaded confidential files.

Issue: Failure to conduct IP diligence on proprietary AI and employee knowledge.

Holding: Court issued injunctions, emphasizing IP enforcement obligations post-employee departure.

Significance: M&A buyers must verify transfer of rights and employee IP exposure.

3. Oracle America v. Rimini Street (2018)

Facts: Rimini Street provided support using Oracle software without proper licensing.

Issue: IP due diligence revealed third-party software misuse.

Holding: Court ruled in favor of Oracle, awarding damages.

Significance: Shows that licensing compliance is a critical part of AI IP diligence, particularly for acquired software-based companies.

4. Facebook v. Power Ventures (2010)

Facts: Power Ventures used Facebook’s data and APIs without permission.

Issue: IP diligence could have revealed licensing and access violations in a potential M&A scenario.

Holding: Courts upheld Facebook’s claims for misuse of confidential information.

Significance: M&A due diligence must assess data access rights, API licenses, and user agreements.

5. Cellectis v. Precision BioSciences (2019)

Facts: Patents for gene-editing AI-assisted methods were contested.

Issue: IP diligence needed to evaluate patent validity and licensing obligations before potential acquisition.

Holding: Courts upheld the patents.

Significance: AI M&A buyers must review patent portfolios for validity, enforceability, and scope.

6. IBM v. Papermaster (2008)

Facts: IBM sued Papermaster for breach of confidentiality after he left to join Apple.

Issue: IP diligence revealed risks of employee knowledge transfer.

Holding: Court enforced NDA and trade secret protection.

Significance: Evaluating employee IP exposure is crucial in AI M&A transactions.

7. Zymergen v. Various Biotech Firms (2020)

Facts: Disputes over ML-assisted synthetic biology processes.

Issue: IP due diligence needed to assess algorithmic patents and licensing.

Holding: Courts upheld patents combining AI and synthetic biology methods.

Significance: AI M&A diligence must review hybrid AI-biotech patents to ensure clear ownership.

*8. Microsoft v. Activision Blizzard (2022–2023)

Facts: Microsoft acquisition of Activision required careful IP due diligence on gaming AI algorithms, player analytics, and proprietary engine code.

Issue: Assessment of trade secrets, licensing, and third-party content rights.

Holding: Pending regulatory approval, but M&A diligence ensured IP was clearly owned/licensable.

Significance: Large-scale AI software acquisitions need comprehensive IP audits, including algorithms, datasets, and licensing compliance.

Key Takeaways for AI M&A IP Due Diligence

Employee IP risk: Ensure NDAs and contracts prevent inadvertent transfer of trade secrets.

Patents and licenses: Verify ownership, enforceability, and freedom-to-operate for AI models and algorithms.

Datasets: Ensure all training datasets are licensed, lawful, and non-infringing.

Open-source compliance: Open-source components require careful review to avoid contamination of proprietary code.

Litigation and dispute history: Review prior IP lawsuits and ongoing claims to mitigate M&A risk.

Documentation: Keep detailed records of IP rights, assignments, and employee agreements.