It should be no surprise to readers of this blog that questions about artificial intelligence (AI) have been the hot topic for the past six months. We have previously written posts related to AI contracting pointers, updates from the US Copyright Office and the UK government, and transatlantic developments, as well as copyright risks. This month, we are going to continue the trend and raise some additional issues related to AI that are relevant to our tech and sourcing followers.
The Usual Suspects
Lawyers and other sourcing professionals that have had the “pleasure” of organizing data rooms or otherwise creating responses to due diligence requests know how burdensome it can be to recreate the inventory of contracts, policies, or other documents that might be responsive to audits, mergers and acquisition (M&A) activity or internal reviews.
While it is true that some technology-related issues are temporal and only briefly find their way into due diligence request lists, others represent significant enough risks to be immortalized in questionnaires, risk assessments, and reps and warranties sections of purchase agreements.
Accordingly, many companies have policies in place to document the usual suspects of such requests as they arise, such as open-source licenses, security incident reports, exclusivity obligations, and non-compete agreements. Identifying these types of documents as they arise greatly reduces the burden of having to retroactively generate and produce them when the need inevitability arises.
A New Face
It is time to add AI related agreements to these lists. Morgan Lewis has been asked by clients to generate reps and warranties that require disclosures of AI-generated work product or other outputs used by the target and to identify any data that has been used to seed or train AI algorithms. Even outside of the M&A space, Morgan Lewis has been asked to include and review provisions that limit or require the disclosure of any AI-related programs that would be used in the creation of any work product in master services agreements.
Currently, we are seeing these types of provisions included in tech-heavy transactions where they are most applicable, but it is reasonable to assume that AI will follow the trajectory of privacy and security and open-source-related provisions and make its way into the boilerplate of agreements across industries.
Accordingly, it may be a good opportunity to take a fresh look at policies related to documenting agreements that are likely to end up the subject matter of diligence requests to ensure that appropriate steps are taken.