Over nearly 20 years in M&A, I have worked in several different team structures. When I compare them with how I work today, I see a clear difference: the path between me, the overall picture and the individual details has become shorter.
I have been involved in transactions of many different kinds and am used to understanding the overall picture, identifying what matters and carrying responsibility for the advice to the client. But with my current way of working, I also have a more coherent and detailed picture of how individual issues move through the transaction.
I am not only in control of the major issues. I can also see how individual findings have arisen, which assumptions they rest on, how they change during the process and where they are ultimately resolved.
The difference does not lie in what I do. It lies in how the work between me and the material is organised.
The traditional transaction pyramid
A traditional legal transaction team is often structured as a pyramid. Junior lawyers go through the data room, review agreements and compile their findings. They report to a senior project manager who holds the work together, sets priorities and decides what needs to be raised with the responsible partner.
It is a rational model. It makes it possible to handle large volumes of material and protects the responsible adviser's time from matters that others judge to be of no significance.
But every handover also involves a judgement. The junior lawyer decides what to raise with the project manager. The project manager decides what needs to reach the responsible adviser. The information becomes more concentrated at each link in the chain, which is usually the whole point.
The consequence, however, is that the person ultimately responsible for the advice often encounters the transaction through several other people's selections. That does not necessarily mean that anything has gone wrong. But the path back to the original detail can become long, and an issue that first seemed insignificant may later turn out to affect something else entirely.
My team looks different
In my current way of working, the internal legal team looks different. I work directly with Yves, my AI colleague, throughout the process.
This does not mean that the entire transaction team consists of the two of us. The client, financial and tax advisers, the counterparty and other people naturally have their roles. The difference lies in how the legal work on my side is organised.
Yves works with the material from due diligence and Q&A through to reporting, agreement versions and closing documents. He can hold together large amounts of information, compare versions and return to earlier questions and instructions. When he summarises a finding, I can go straight to the source, test the reasoning and follow the issue onwards through the process.
That does not make the information unfiltered. Yves also makes judgements, and I correct them when they are too categorical, miss the commercial point or give a detail more weight than it deserves. The difference is that the judgement and the source sit close to each other. I do not need to ask several people to reconstruct how the conclusion took shape.
An issue moves through the entire transaction
The real value becomes clear when an issue changes shape over the course of the transaction.
A provision in an important customer contract may at first appear to be one of many findings in the due diligence. When it is followed through Q&A, the picture becomes more nuanced. The same issue then needs to be weighed against the company's forecast, the buyer's risk appetite and the contractual mechanics. Finally, it may need to be dealt with through a customer consent before closing, a warranty in the purchase agreement or a condition that gives the buyer the right to withdraw if the situation changes.
In the same way, a wording in a new draft agreement can be understood against the entire negotiation history. Who introduced it? What problem is it intended to solve? Have the circumstances changed since then? It is easy to react to the words in the latest version. It is harder, but more important, to understand why they are there.
The work of other advisers also needs to be connected to the same overall picture. A financial or tax finding may affect the price adjustment, the definition of debt or the design of a specific indemnity mechanism. If the different workstreams are read separately, there is a risk that the same circumstance is counted twice or that it falls between two areas of responsibility.
In the final stage, the connections continue. An obligation in the purchase agreement needs to be matched by the right closing document, the right signatory and the right timing. At that point, it is not enough to know the final version of the agreement. You also need to understand why the requirement exists and what happens if the document is missing.
It is this movement through the transaction that I can now follow with a different closeness than before. Not just document by document, but from the first finding to the way the issue is finally handled.
My work is still mine
I work the same way I always have. I challenge conclusions, question assumptions, ask for simpler explanations, remove what is not relevant and weigh legal risks against the real circumstances of the transaction. I negotiate, set priorities and take responsibility for the advice to the client.
Yves can read widely, remember connections and quickly produce supporting material. But he cannot decide what the client should accept to get the deal done. He cannot sense when a counterparty needs to be pushed and when an issue should be played down. He cannot carry the responsibility for ensuring that legally sound advice is also commercially wise.
That is still my work.
What has changed is that I can exercise my judgement on a richer basis and with a shorter path back to the details.
Closer to the transaction, not further from it
The discussion about AI in law is often about which tasks the technology can take over. My experience of this way of working has raised a different and more interesting question for me: How should a team be organised so that the responsible adviser understands more of what is actually happening in the transaction?
The answer is not necessarily fewer people in every situation. Large and specialised transactions will continue to require several areas of expertise and many hands. But it is worth considering how many links in the information chain are really needed, how easy it is to go back to the source and whether the responsible adviser sees the connections between the workstreams.
In this way of working, AI does not make my judgement less important. It enables me to apply it closer to the material, more often and across a larger part of the process.
Quite simply, I know the transactions better.