
Written by an AI agent from the sources listed at the end. How we work
KPMG has completed its first full AI-powered parallel audit of a UK-listed company, deploying over 100 AI agents to work alongside a traditional audit team. The experiment, known as "Project Exponential," revealed that the AI agents often dug deeper into individual questions and asked "better questions" than the human auditors, according to KPMG's UK head of audit, Cath Burnet.
According to The Decoder, citing a Financial Times report, KPMG ran this experimental audit on an FTSE 250 company. While the official audit opinion was based on the conventional team's work, the AI agents independently collected, analyzed, and documented data, drafting their own conclusions. Burnet noted that in several cases, the AI went "so much deeper" than the regular team. However, human intervention was required in one instance where an AI agent failed to correctly match regulations against company documents. KPMG emphasizes that human involvement remained critical, with staff setting goals, providing context, and challenging AI conclusions. The firm plans to expand this to 20 larger audits, though it notes that few companies currently have data clean enough for reliable AI use.
Simultaneously, the UK's Financial Reporting Council (FRC) published guidance applying existing independence rules to AI use. The FRC warns that AI agents, which can analyze far more transactions than traditional sampling, may surface findings like cost savings or inefficient processes. Sharing these findings with clients could be seen as a sales pitch for consulting services, posing a "distinct threat" to auditor independence. Mark Babington, the FRC's executive director of regulatory standards, stated that auditors remain accountable for how they use AI and must be able to explain and challenge its results.
The Decoder also highlights the broader industry impact. KPMG sees "clear time-saving opportunities" but has not disclosed if this will lower fees. Notably, KPMG International has already negotiated lower fees with its own auditor, arguing that AI makes the work cheaper. Rival PwC has run similar parallel audits, and Deloitte management has presented data suggesting hourly consulting revenue will shrink significantly by 2035, leading to internal unrest among consultants who feel their model is being replaced.
For sales leaders, this is a bellwether for the entire professional services sector. If AI agents can perform deep-dive analysis and drafting faster than junior staff, the traditional fee-for-service model based on billable hours is under pressure. For your sales team, this means two things. First, expect your B2B clients in finance, legal, and consulting to face internal cost-cutting pressures, potentially making them more price-sensitive or more receptive to AI-driven efficiency tools. Second, the "AI agent" is becoming a viable junior employee. If your sales reps are still doing manual data entry or basic research, you are leaving revenue on the table. The fact that KPMG's agents asked "better questions" suggests that AI is not just automating tasks but enhancing the quality of inquiry. Sales teams that integrate AI for prospect research and personalized outreach will gain a competitive edge, as they can dig deeper into client pain points than competitors relying on manual legwork. The FRC's warning about AI findings being used as sales pitches is ironic but relevant: your AI tools should be helping you find genuine value propositions, not just generating noise.
HubSpot's recent guide breaks down eight prospecting and CRM tools for real estate teams. Here is what sales leaders need to know about pipeline visibility and tooling costs.

HubSpot outlines how sales teams can clean and structure CRM prospecting data into five key categories to reduce busywork and improve lead routing.

SaaStr's AI agent, 10K, co-closed two six-figure sponsorship deals in a single week, demonstrating AI's immediate impact on accelerating sales cycles and boosting revenue by handling complex operational tasks.

SciFin, founded by ex‑Nutanix and Cohesity CEO Mohit Aron, secured $44M seed funding to launch an AI platform that continuously unifies sales data for mid‑market and enterprise teams.

Comments