Marketplace Risk Global Summit (MRGS) is one of the world’s largest digital risk conferences, where trust and safety leaders and platform founders share the strategies they rely on to protect marketplaces and fintechs. The 2026 edition takes place on November 2nd and 3rd in London.

This year’s agenda covers fraud prevention, regulatory compliance, and the technology platforms use to defend themselves as new threats emerge. Across those topics runs one shared question: what happens to real people after a platform makes a decision? A flagged seller, a buyer who meets a scam, a new member stuck at sign-up. Each of them experiences the platform through the outcome of a decision.

Ailís Daly, Head of Trust and Safety at WebPurify, an IntouchCX company, will be in London to bring the team’s hands-on perspective to these conversations and share her expertise at the conference Beyond the Screen: Building Safer Communities and Real-World Connections with Strava, along with Natasha Ahmed Walsh, Head of Public Policy for the UK and Europe at Strava.

Below are three themes on the agenda and what IntouchCX and WebPurify have seen working on them alongside platform teams.

Fraud and deepfakes: when a strong model still leaves members exposed

A platform can have the best detection rates its fraud model has ever produced and still receive user reports about the same scams, while honest sellers lose their accounts to flags that should never have been raised.

The gap exists because a detection rate describes how well the model sorts cases. Members only experience what happens to them after each decision: a wrongly flagged seller loses income and loyalty, and a missed scam reaches a buyer who trusted the platform to keep them safe. A February 2026 press release from the European Commission reports that platforms reversed 30% of the 165 million moderation decisions users appealed under the European Union’s Digital Services Act. Independent dispute bodies overturned platform decisions in 52% of the cases they closed in the first half of 2025. Each reversal represents a member who was wrongly stopped.

Identity checks face the same pressure, and AI is raising the stakes: Forged payslips and AI-generated documents now pass checks that once caught them by eye, while every added verification step adds friction to a business that depends on conversion. IntouchCX saw this firsthand in a driver identity verification program built for a global mass-market retailer, where drivers had to clear a review of their driver’s license and a selfie. While validating those documents, the team found fraud patterns in active use: AI-generated licenses sharing identical signatures across multiple submissions, AI-generated live video standing in for real-time liveness checks, altered state licenses with computerized typeface signatures, and AI-generated state IDs paired with AI-generated profile photos.

The program treated verification as a fraud intelligence function embedded in onboarding, keeping 99% quality and an average of 47 seconds per profile. Fraudulent users were caught, and legitimate drivers were not slowed down.

These are the patterns MRGS sessions on deepfakes and fraud-as-a-service will dig into: how AI-generated faces get past facial biometric checks, why detection alone cannot keep pace, and how tying identity to the physical world builds a stronger defense. The platforms making real progress measure what happens after the decision, asking whether harmful behavior declined and whether legitimate members stayed protected.

Product safety: seeing the problem before the regulator does

The same logic applies when the person affected is a buyer, and the decision is about what stays listed. A marketplace selling thousands of products across several countries can receive a safety alert about one product category in a single jurisdiction, and by the time that alert reaches the review queue, the same listing may already have sold in other markets.

Rules such as the Digital Services Act, which sets timelines for handling notices, appeals, and reports from trusted flaggers, raise the bar for how quickly platforms respond. Enforcement agencies also set priorities of their own, so a platform watching only its internal risk signals can miss where regulators are about to focus. Visibility across jurisdictions has become part of a compliance team’s daily work.

Regulators are now penalizing platforms that wait for harm to surface: In July 2026, the European Commission fined AliExpress €550 million under the Digital Services Act for failing to diligently assess the risk of illegal, unsafe, or counterfeit products on its marketplace and to act effectively against it. The Commission made clear that these risks must be identified and addressed systematically, which means looking for problems before a notice arrives.

At MRGS, sessions on product safety enforcement will cover how agencies prioritize action under the Digital Services Act and the General Product Safety Regulation. From WebPurify’S experience, the teams that stay ahead keep each procedure tied to its requirement, so a new rule or a shift in regulator focus becomes an edit to an existing process instead of a scramble.

Authentication: the SMS passcode deserves a second look

Many platforms treat the one-time passcode sent by text message as a fixed cost of doing business, expensive and occasionally frustrating yet seen as necessary. For members, it can mean a delayed code, a failed sign-up, or an account exposed to SIM-swap attacks.

Some teams are now testing that assumption by replacing the code with passive, key-based authentication, which confirms a member’s device in the background and asks for nothing to type or share. A 2026 SANS Institute study performed in a controlled lab environment built on Microsoft Entra ID, found that SMS one-time passcodes were “highly vulnerable” to automated phishing kits that sit between the user and the real website. It also found that enforcing FIDO2 security keys, which only respond to the legitimate site, “completely blocks token hijacking, reducing bypass rates to absolute zero”. The change from sms authentication to security keys touches metrics owned by several teams at once: fraud exposure for risk leaders, operating cost for finance and engineering, and sign-up conversion for product.

The summit will cover the technical and organizational case for the switch and what early adopters have learned so far. IntouchCX supports platforms on identity verification and account security, where every check trades some friction for some protection. The useful question for any team heading to London: is the authentication stack quietly working against the members it was built to protect?

See you in London

A wrongly flagged seller, a buyer who receives an unsafe product, and a member locked out by a text message code all share a starting point: a platform decision that lands on a real person. Regulators now expect platforms to find those problems before customers report them, and MRGS is where the industry compares notes on how.

Ailís Daly and the IntouchCX team will be at Marketplace Risk Global Summit 2026 on November 2nd and 3rd.

Join us to compare, share, challenge, and evolve the practices shaping trust and safety today.