Insuring Fraud at Signup: Sunil Madhu on AI Underwriting, Faster Approvals, and Claims in 30 Days

Send us a text

Title: Insuring Fraud at Signup: Sunil Madhu on AI Underwriting, Faster Approvals, and Claims in 30 Days
Guest: Sunil Madhu, Founder & CEO, Instnt (previously Founder & CEO, Socure)
Length: ~41 minutes

What This Episode Is About

Most fraud tools try to detect risk. Instnt goes a step further: it underwrites onboarding fraud risk in real time and transfers residual loss to A-rated insurers. For consumers, that can mean fewer false declines and faster approvals. For institutions, it can free up risk capital and reduce the pressure to add friction “just in case.”

Who Should Listen

• Banking, fintech, and payments leaders balancing growth vs. fraud
• Compliance and risk teams navigating KYC/AML, Reg E, and UDAAP
• Curious listeners who want a plain-English look at how onboarding actually works

Key Topics We Cover

• The customer journey: where the binary “insurable / not insurable” decision fires during signup and what changes for the user experience
• What’s novel: pairing AI risk assessment with insurance capacity—not just a score, but loss transfer
• Claims & operations: online filing and marketed ~30-day payouts; how coverage coexists with chargebacks and statutory consumer redress
• Risk incentives: preventing moral hazard and keeping approvals smart (model governance, recalibration, bias checks)
• Compliance fit: how the model sits alongside existing KYC/AML stacks and why it doesn’t change Reg E obligations to consumers
• Market impact: where traditional defenses fail (e.g., synthetics, first-party fraud) and how insurability changes economics

Plain-English Glossary

• Onboarding fraud: Fraud that happens while opening or activating an account (e.g., stolen or synthetic identity).
• Loss transfer: Moving expected fraud losses from the institution’s P&L to an insurance policy.
• False decline: A legitimate customer wrongly blocked or forced through excessive friction.
• Moral hazard: The risk of loosening controls just because losses are insured.

Takeaways

1. Consumers feel it: Insuring residual fraud risk lets institutions remove friction and cut false declines, improving first impressions.
2. Finance cares: Moving expected losses off the books can free reserves, turning fraud spend from a pure cost to a growth enabler.
3. Policy still binds: Insurance doesn’t erase Reg E or consumer redress—coverage is about the institution’s residual loss, not limiting statutory rights.
4. AI with accountability: Real-time underwriting must come with explainability and recalibration to avoid bias and drift.

Suggested Chapter Guide (approx.)

• 00:00–04:30 | Why fraud insurance at onboarding exists
• 04:30–12:00 | The “apply → approved” flow and the insurability decision
• 12:00–20:00 | Claims, payouts, and what changes for ops teams
• 20:00–30:00 | Compliance fit: KYC/AML, Reg E, data privacy
• 30:00–40:00 | Incentives, bias, and the future of AI + insurance

Resources

• Guest: Sunil Madhu (LinkedIn)
• Company: Instnt — identity fraud loss insurance for onboarding (AI underwriting + A-rated insurer backing)

More from Jim:
LinkedIn: https://www.linkedin.com/in/mccarthyhatch/
https://www.mccarthy-hatch.com/