You’re buying an asset – and it includes AI. What does that mean for your due diligence approach? AI concerns are a major cause of M&A deals collapsing, with security and compliance seen as the biggest risk areas¹. So, are the old levels of vetting and risk assessment no longer enough?

Well, ask someone who’s done it. Datasite’s recent acquisitions have included Sourcescrub, Grata, and Blueflame AI – all based around artificial intelligence. And with home-grown AI embedded in the platform for years, Datasite is also the first data room provider to achieve ISO/IEC 42001 certification for responsible AI governance. With that in mind, here are our top tips for approaching any deal that involves AI assets.

1. Data security

Examine how the target’s AI technology addresses data security issues. In particular, evaluate:

1. Encryption protocols and access controls.
2. Mechanisms for preventing unauthorized access.
3. Compliance with relevant data protection regulations.

Rationale

Aside from the obvious – preventing data breaches and compliance failures – it is vital to ensure that the target’s AI meets the security standards of the acquiring company. Otherwise the acquired technology is itself a weak point, reducing security standards for the entire entity.

2. Human workflows

Look at how well the AI technology aligns with how people currently work. Most importantly, consider:

1. The user interface. Is it friendly, straightforward, and easy to master?
2. Existing systems. How well will the new technology integrate with them?
3. The changeover. Will it require much training and/or change management?

Rationale

‘If it ain’t broke, don’t fix it’ is a strong sentiment in every office. Change that assists or enhances workflows will be embraced – but even small obstacles generate resistance and the sense that ‘things were better before’. Be alert to potential friction.

A good example of alignment with existing systems can be found in Datasite’s acquisition of Grata. Since Grata sources data from private companies in the mid-market, it integrated extremely well into Datasite’s existing app Pipeline, for identifying and managing targets. Now users can pull in live data to their pipelines – enhancing their experience with no need to switch platforms.

Inspect the intellectual property rights associated with the AI technology itself. Pay special attention to:

1. Verifying ownership and protection of the technology.
2. Assessing the risk of infringement claims.
3. Reviewing licensing agreements and patents.

Rationale

Your AI should be an asset, not a hidden liability. Beware of buying up technology that makes you a litigation target. Also ensure that the asset represents value for money – which it won’t if the license or patent expires too soon.