The rise of AI-generated work for legal teams
Ask most legal leaders what they expect from AI and you'll typically hear the same requirements: less admin, faster turnaround and more time for the work that actually matters.
The reality so far has been mixed.
From our experience, AI does speed up some of the work legal teams already do, but it's also creating high volumes of new work: more contract mark-ups, data subject access requests (DSARS), complaints and queries arriving in lawyers' inboxes, because AI has made it easier for everyone to formulate a legal question.
The key question for legal teams is a practical one: what needs to change in how we run things, so we can handle higher volumes of work within the current team?
That's fundamentally a legal operations question rather than a technology one. As we say in Rising Roles in Law:
in any walk of life, a team of superstars needs glue to support the structure. Legal operations is that glue.
Start by deciding what doesn't need a lawyer.
More AI-generated work means more matters landing on a legal team's desk without warning.
That might be a commercial contract drafted at speed, or an employment query or consumer complaint that was produced by an individual in minutes with the help of AI.
The answer to dealing with this new influx is not always more headcount. More often than not, it's about better sorting so fewer requests reach a lawyer in the first place.
In practice, that means splitting work three ways:
- Low-risk, low-value matters remain with the business through self-service processes with clear guardrails.
- Strategic work stays with the in-house team, where knowledge of the business matters most.
- High-volume, routine work gets routed to alternative providers rather than absorbed internally.
A single, structured intake point (or "Legal Front Door") can do that sorting before anything lands on a lawyer's desk. In practice, this could be a short intake form with a few trigger questions built in, which routes requests automatically based on type, risk and/or value.
This does not have to mean new technology – plenty of legal teams use workflow tools their company already has available, repurposing them as legal intake and matter management platforms.
Then, create AI-compatible guidance materials.
Getting the quality of AI outputs right increasingly depends on the knowledge sitting behind the tool itself – not just on checking its answers afterwards.
Much of the useful information may already sit inside the legal team, such as:
- precedents;
- policies;
- playbooks;
- previously drafted advice;
- escalation principles;
- worked examples of how the business usually approaches a risk.
The task is to turn that material into something an AI tool can use safely.
It needs to be structured, current, de-duplicated, cleansed and clear about which points are guidance rather than authority.
This is where having a lawyer check the materials before go-live matters most. Anything intended for repeat use, such as an AI-assisted negotiation guide, an approvals pathway or an automated response, should have clear checkpoints built into the workflow. If the material needs legal judgement, a lawyer should review it before it goes further.
In practice, materials created for AI ingestion should be treated like any other new legal process. Test them with the people who will actually use them, refine them, and only then rely on them.
Finally, adoption itself is delivery work.
In Lewis Silkin's Future @ Work 2026 report, nearly half of organisations surveyed said they expect cultural resistance to slow the adoption of new technologies and operating models. That will feel familiar to many in-house teams. The tools may work, but the people around them often need help to adapt. New workflows need proper preparation. That means:
- holding discovery sessions with the people affected;
- understanding what is actually breaking in their day-to-day process;
- identifying what needs technology vs better self-serve resources vs a slicker process;
- creating materials that reflect how people actually work.
Teams that skip this readiness phase often find that the technology performs well, but adoption still does not happen.
As we note in Rising Roles in Law:
the ability to apply an empathetic lens to how changes in a process will impact the people involved is something AI has yet to mimic.
And it's precisely that human judgement that makes readiness work effective.
None of this is a reason to slow down. It is a reason to stop treating AI as a procurement decision with a fixed implementation date.
AI needs to be treated as ongoing delivery work, with intake, checking, quality control and adoption all built and maintained over time. AI – the same force creating the pressure – can also help legal teams respond to it, provided the operational infrastructure is in place.
This shift is already visible in how legal services are being delivered. Across commercial, privacy and employment work, people are asking for hands-on implementation support, not just recommendations.
The legal teams getting real value from AI are not always the ones with the flashiest tools; they are the ones that worked out, early, that more AI means more to manage, and then put the structure in place to manage it well.
This article draws on themes explored in more detail in Rising Roles in Law: Designing the AI-Enabled, Data-Driven Law Firm (Globe Law and Business, 2026). If your team is navigating any of the challenges above – whether that's building an intake process, structuring guidance materials for AI, or working out where to start with legal operations – we'd welcome a conversation. Get in touch with us or find the book here.
