Sep 18, 2026 · 51 min
Accounting firms confront AI, fraud and a broken client experience
Feds Target KPMG Polymarket Insider Trading, ComplYant Founder Charged with Wire Fraud
The episode connects emerging AI risks and accounting scandals to a more basic challenge: firms must remove the work and friction clients already pay them to handle.
- 1Polymarket bets linked to KPMG-audited companies raise questions about insider information and prediction markets’ role in detecting abuse.
- 2KPMG’s AI efforts span tax training, enterprise implementation and deepfake detection as firms adapt to automation and fraud.
- 3The hosts argue accounting firms should pass efficiency gains to clients and redesign services around seamless execution rather than hourly effort.
Don't miss
The hosts turn their AI tax-return experiment into the episode’s clearest diagnosis: accounting firms should make clients’ lives easier instead of outsourcing administrative work back to them.
The brief
Blake Oliver and David Lirian open with a challenge to accounting firms: fearing AI misses the larger problem when clients still manage much of the work they pay professionals to handle.
Anonymous Polymarket accounts with unusually strong bets on companies audited by KPMG prompt questions about insider information, blockchain transparency and whether prediction markets can expose market abuse.
KPMG’s response to a changing profession includes TaxSim training, an OpenAI alliance and investment in Reality Defender, linking accountant education with the growing business of AI-enabled fraud.
The hosts examine alleged investor fraud at tax startup Compliant, then revisit Xero’s apology after an influencer suggested business owners could replace accounting services with software and Claude.
After testing AI alongside TaxAct and TurboTax on S corporation returns, they argue firms should focus less on defending expertise and more on eliminating client back-and-forth.
IRS guidance on passing efficiency savings to clients sharpens the episode’s central tension: professional responsibility is difficult to reconcile with billing systems built around time spent.