At the recent finance leadership panel hosted by Light at Amazon Web Services (AWS)'s Amsterdam office, CFOs and investors explored "How AI Flipped the Script of Financial Reporting." The discussion was lively and eye-opening: can AI finally deliver the holy grail of real-time financial reporting, and do we even want it if it could?
The event started with a demo of Light's soon-to-be-released text-to-chart function:
Real-time reporting – boon or bane? Panelists were divided on whether instant "close-free" reporting is desirable. "In general I'd be in favor, but I hesitate…you need to make sure the data is actually correct," noted Werner Huijboom (founder, Soigné), emphasizing that finance teams won't embrace real-time data without trust in its accuracy. The group agreed that speed means nothing without reliability. Uvashni Raman (CFO, Booking.com) stressed that real-time only works if built on a solid infrastructure and controls:
“I'm all for giving the right information to the right people as quick as possible…There have to be checks and balances in place…so you're confident the information you're getting is right”
In other words, faster reporting is great – but not at the expense of accuracy and context.
Who do you trust – AI or human? This question sparked debate about trust and accountability when AI enters finance. "Would you trust an instant AI-generated report?" the moderator challenged. The consensus: trust has to be earned. Light's co-founder Jonathan Sanders (who earlier demoed Light's AI-driven finance platform) explained that building confidence in AI starts with design and transparency. For example, AI can flag whether figures come from a closed period vs. open books, check that all expected transactions are included, and even trace each number back to its source document.
“Ideally, with AI you should be able to pinpoint exactly which document a figure came from, and where in the document”
Jonathan noted, highlighting how such transparency can actually increase trust in the numbers.
Yet, panelists acknowledged a human element will always remain. As Uvashni put it, "It doesn't take away the fact that you still have to rationalize whatever the output is…The human judgement element will always be there." Jeyne Chun (SVP Finance, Adyen) agreed, emphasizing that CFOs are still on the hook:
“Accountability doesn't change…If the CFO is responsible for financial reporting, the accuracy of the financials is [still] on the CFO.”
AI or not, finance leaders won't be off the hot seat – but AI can help them be confident that what they're signing off on is accurate.
From number-cruncher to strategic partner. Perhaps the most exciting takeaway was how AI can free finance teams from grunt work. "Ideally, AI allows you to get rid of the boring stuff and focus on the more important, strategic questions," said Alex A. (VC, Northzone). Several panelists echoed this point: rather than spending days chasing missing receipts or tweaking spreadsheets at month-end, teams can redirect energy to analysis and strategy. One memorable example, described by Jonathan:
“Our AI-finance agent will automatically ping employees for missing expense receipts and context (like why a team dinner cost €1,500) – saving the CFO from being the 'bad cop' on Slack.”
See the Light Agent act as "bad cop" here.
Removing that friction makes compliance a cooperative process and leaves humans to handle exceptions and big-picture questions. As Alexander quipped, "the finance function should be as lean as possible…with amazing professionals working on strategic questions, and all the boring stuff cut out by AI."
