So why are 45% of those same firms telling us they've created more work when it comes to validating AI-generated output?
On the surface, that seems contradictory. If AI is saving time, why are firms still feeling pressure on resources? The answer is more nuanced than simply measuring time saved on a task.
The efficiency story has a gap
The investment picture looks positive. Nearly all firms (96%) are now using AI within support services, and more than half (56%) have hired dedicated AI specialists to support adoption efforts. Firms are making investments, and they're moving quickly.
But the data also tells us that 45% of firms are spending additional time reviewing and verifying AI outputs. In many cases, the time saved at the front end of a process is being offset by work that's happening further downstream.
That's not necessarily a problem. Quality control is extremely important. The challenge is when firms haven't thought through who should be doing that work, how it should be managed, or how much time it's actually consuming.
If you don't have visibility into where work is happening and who is doing it, it's difficult to understand whether you're truly gaining efficiency or simply shifting effort from one place to another.
Adding new technology to outdated operating models
This is where I see many firms getting stuck.
Technology is moving fast, but the operational changes needed to support it are moving at a snail's pace.
According to our research:
- Only 41% of firms have completed a formal review of support task types to identify where automation will deliver the greatest value.
- Just 27% have redesigned support roles or operational structures to reflect how work needs to flow in an AI-enabled environment.
What that tells me is that many firms are introducing AI into operating models that were never designed to support AI-enabled workflows, or the way work needs to flow across a modern support function.
Technology can absolutely improve efficiency. But if the underlying processes, workflows, and support structures haven't evolved alongside it, firms often fail to realize the full value of their investment.
Thomson Reuters reached a similar conclusion in its 2025 Future of Professionals Report, finding that organizations with a clear AI strategy are 3.5 times more likely to achieve a return on their AI investment. The difference isn't simply technology adoption. It's having a deliberate plan for how technology, people, and processes work together.
The governance challenge firms still need to address
There's another issue that deserves more attention: accountability.
As AI becomes more embedded in day-to-day support work, the need for governance grows with it. Someone needs to be responsible for validating outputs, monitoring quality, and ensuring standards are consistently applied.
In my view, it shouldn't sit primarily with lawyers, whose time is best spent focused on client work. Nor should it fall disproportionately on experienced support professionals, whose greatest value comes from relationship management, coordination, and applying their experience and judgment where it has the greatest impact.
This is where centralized support functions become increasingly important.
A well-structured central support model creates clear ownership, consistent processes, and better visibility across the firm. It provides a place for governance to live and for accountability to be clearly defined.
The reality is that many firms haven't yet built that structure, which means responsibility for AI oversight is often spread across multiple teams with no clear owner.
What operational readiness actually looks like
The good news is that firms don't need to slow down their AI initiatives. They do, however, need to build the operational foundations that allow those initiatives to succeed.
In practical terms, that means:
- Structuring support functions so work is consistently routed to the right resource, at the right level, at the right time.
- Ensuring staff receive the training and upskilling required to effectively review and validate AI outputs.
- Identifying opportunities for process redesign before introducing new technology, rather than after.
- Creating the visibility needed to understand where work is being completed, where bottlenecks exist, and where AI is genuinely delivering value.
Technology is only part of the equation. Sustainable efficiency comes from understanding how work moves through the organization and making sure the right people are focused on the right activities.
This is exactly the type of challenge workflow technology is designed to solve. AI Email Routing within BigHand's Workflow Management solution automatically reads emails submitted to a centralized support inbox, identifies the work required, and routes it to the most appropriate resource. AI handles intake and classification, allowing support professionals to focus on the work that requires their expertise.
That's what effective AI adoption looks like. Not simply layering technology onto existing processes, but removing friction, improving visibility, and creating a structure that allows people and technology to work together effectively.
The firms that achieve lasting value from AI will be the ones that invest as heavily in their operating model as they do in the technology. AI may be advancing rapidly, but long-term success will depend on whether firms are building the structures needed to support it.
If you're evaluating how AI fits into your firm's broader operational strategy, I encourage you to download the BigHand Legal Workflow Leadership Report. The data provides valuable insight into where firms are investing, where they're seeing results, and where challenges still remain.
