AI was supposed to take manual work like copying data between systems, reconciling discrepancies and chasing receipts off finance’s plate. While AI has reduced some of that manual work, it hasn’t eliminated those tasks (despite the enthusiastic predictions). In fact, 50% of finance teams* say their finance tech stack is still manual, according to new research from Pleo, and the reason comes down to disconnection. Teams have added AI tools without connecting them to the rest of the stack, so the manual work just moves rather than disappears.
Despite incremental productivity gains, finance teams shouldn’t lose faith in AI’s ability to reduce manual tasks. AI works best once finance addresses the following factors.
One teammate swears by Claude and another uses ChatGPT, all while teams switch between email, Slack, accounting platforms and other systems to complete everyday tasks. It’s a common problem among finance teams. In fact, 47% say they’ve over-invested in different AI tools and are now dealing with AI sprawl, according to Pleo research*. Teams end up with separate tools, logins and exports, with someone in the middle still piecing it all together by hand.
Despite this AI sprawl, it’s unrealistic to use one AI tool exclusively. The tools have specific strengths and pricing structures. Instead, configure your tech stack to bridge the AI tools finance already uses, rather than asking the team to learn another one.
A connected tech stack lets teams take action and run workflows from inside LLMs like Claude and ChatGPT, or other platforms like NetSuite and Microsoft 365. Use the systems already on a team member’s computer, rather than having them toggle between multiple tools. That connection removes a layer of manual admin instead of adding to it, turning AI tool adoption into time saved, not just another subscription to manage.
AI can provide answers around spend, budgets or approvals within seconds. Even where AI is genuinely useful, there’s often a gap between the insight and the action. A finance team might ask an AI tool to pull spend data or flag an anomaly, get a clear answer and then still have to manually make a correction. The fix doesn’t happen on its own.
That’s where the manual work creeps back in.
The shift to actually reducing manual work happens when repetitive, multi-step expense tasks, such as chasing a missing receipt, coding a batch of transactions or flagging what’s outside policy, become a single task. Then take it a step further by creating automated workflows across the tools finance already relies on, such as Slack, NetSuite or Microsoft 365.
Closing that gap shouldn’t mean waiting on another team to build the fix. Give your finance team members the autonomy to create workflows that suit their process without calling on engineering to build the steps.
Finance leaders often can’t fully see how AI tools reached an answer. When they can’t see the logic, it’s harder to trust the output. So teams spend unnecessary time manually cross-checking information and chasing paper trails just to ensure 100% accuracy.
To maintain confidence and reduce time wasted double-checking information, your finance tech stack should show a full audit trail on every action, require human direction during specific scenarios and prevent certain users from making changes or updates to workflows without permission.
AI earns its keep when it knows exactly what it should handle on its own and where it needs to step back (scoped to real permissions and logged clearly) so nothing happens without you knowing why.
None of this means finance needs to rethink its AI strategy from scratch. Instead, focus on connecting the tools already in use, so answers, actions and workflows all run from one place, instead of from scattered reports and dashboards.
Get that right, and finance stops manually switching between systems and starts spending its time forecasting, shaping strategy and guiding where the business should invest next.
*This research is taken from a forthcoming report from Pleo, launching in September 2026. The survey was conducted by Sapio Research in April 2026, among finance decision-makers and finance professionals in organisations with 51 to 1,000 employees in Germany, Spain and the UK.