How is the increase in AI assets going to impact Intellectual Property due diligence in M&A transactions?

Well the truth is, who knows… But with OpenAI, Google, Microsoft and Amazon Web Services (AWS) rolling out AI products that are enabling companies to build their own custom AI systems and applications, alongside publishing content developed by AI, It can be argued that M&A transactions which involve ‘AI assets’ will become the norm, and not the exception in the near future.

Our CEO, Mark Purdy, has been speaking with some of The Code Registry‘s customers in the PE, VC, M&A space on this topic recently and it’s something that is cropping up more and more in the tech due diligence space. Some of the key takeaways for both parties during an M&A transaction we’ve discussed were;

Ownership and Liabilities

AI introduces complex intellectual property (IP) issues, such as determining ownership and liability for AI-created assets, which can vary by country. This is especially prevalent within software, so business leaders need to be aware of what they have ‘under-the-hood’ within their software assets and what they can claim as their own IP. Ownership of AI-generated content and algorithms can be complex, often involving multiple creators and proprietary technologies. Determining clear ownership rights is crucial to avoid future disputes and ensure that the acquiring company fully understands what they are purchasing. Additionally, liabilities associated with AI assets must be carefully evaluated. This includes potential infringements on third-party IP, compliance with data privacy regulations, and any existing litigation related to the AI technologies. Thorough due diligence in these areas is essential to mitigate risks and ensure a smooth transition during M&A activities.

Identification of Assets

Companies must be able to identify all open-source or external components of their software, including AI assets such as algorithms, training data, and code, as these can be subject to different IP rights.

A comprehensive inventory of these assets is critical during IP due diligence in M&A transactions. This involves not only recognizing the proprietary elements but also understanding the dependencies and integrations with third-party technologies. Proper identification helps in assessing the full scope of IP ownership and any associated risks.

Furthermore, special attention should be given to the provenance and usage rights of training data used in AI models. Ensuring that the data was acquired and used in compliance with applicable laws and agreements is essential to avoid potential legal complications post-acquisition. Companies need to evaluate the licensing agreements associated with open-source components to confirm compliance and identify any obligations or restrictions that might affect the future use and development of the software. This meticulous process of identifying and cataloging AI assets ensures that the acquiring company has a clear understanding of the IP landscape and can make informed decisions, minimizing unforeseen liabilities and enhancing the value derived from the transaction.

Open Source Licences

With the rise of pre-trained open-source AI models, understanding the license terms and potential restrictions on AI asset usage is more critical than ever. Open-source licenses can vary widely, from permissive licenses that allow extensive freedom to use, modify, and distribute the software, to more restrictive licenses that impose certain conditions, such as copyleft clauses that require derivative works to also be open-source. During IP due diligence in M&A transactions, it is imperative to thoroughly review these licenses to ensure compliance and avoid future legal issues.

Compliance with legislation and regulations is also essential for both buyers and sellers in M&A deals. This includes adhering to data protection laws, especially when dealing with AI assets that involve personal or sensitive data. Regulatory compliance extends to ensuring that the use of open-source AI models does not violate export control laws or industry-specific regulations. For instance, certain jurisdictions may have stringent rules governing the deployment and use of AI technologies in healthcare, finance, or defence sectors.

By diligently assessing open-source licenses and regulatory requirements, companies can mitigate the risk of legal disputes and financial penalties post-acquisition. This due diligence not only safeguards the transaction but also ensures that the AI assets can be seamlessly integrated and utilized within the acquiring company’s operations, maximizing the strategic value of the M&A deal.

Protection Measures

Protecting AI assets, just as you would protect other digital assets as trade secrets, is vital, especially when patent or copyright protection isn’t available. AI algorithms, models, and proprietary datasets can be highly valuable and competitive assets. Implementing robust protection measures ensures that these critical resources are safeguarded from unauthorized access and potential theft.

Evaluating your code, security, and IT infrastructure is a critical component of this protective strategy. This involves conducting regular security audits to identify and address vulnerabilities, ensuring that access to sensitive AI assets is restricted and monitored. Encryption, secure coding practices, and stringent access controls should be standard to protect the integrity and confidentiality of AI technologies.

Additionally, it’s important to establish comprehensive internal policies and procedures for handling AI assets in the same way you would any other digital asset. This includes clear guidelines on data management, employee training on security best practices, and protocols for responding to potential security breaches. By proactively protecting AI assets, companies not only preserve their competitive advantage but also enhance their attractiveness to potential buyers in M&A transactions. Strong protection measures demonstrate a commitment to safeguarding intellectual property, thereby instilling confidence in the transaction’s long-term success.