AI-led fraud intervention

In the run-up to a high-profile product launch, this leading telecoms provider faced a sharp spike in fraud risk, both first-party and third-party, that threatened to derail its performance. With just weeks to act, the business partnered with Jaywing to deploy AI-led defences at pace and scale.

Jaywing used its proprietary AI modelling platform, Archetype, to build and deploy a suite of targeted fraud models, capable of identifying and blocking fraudulent activity in near real-time. These models were built in under four weeks and calibrated to avoid unnecessary friction for legitimate customers.

The results were immediate:

- c.70% reduction in third-party fraud
- c.20%+ reduction in first-party fraud
- Tens of millions in savings, outperforming the original target
- Live model updates delivered during the iPhone launch period

This was not AI for show. It was a live, embedded part of the telecoms business’s operational risk framework, delivering measurable commercial protection. Jaywing also transferred skills, model explainability tooling, and governance structures to ensure the firm could maintain and evolve the solution in-house.

By combining modelling rigour with real-world urgency, Jaywing delivered one of the most effective AI-led fraud interventions in the telecoms sector — protecting customers, reducing losses, and setting a new standard for managing fraud risk in high-pressure environments.