Britain and its Five Eyes partners are pushing fraud prevention further into the infrastructure used to deliver scams, with governments agreeing to deepen cooperation between law enforcement, technology companies, communications providers, and financial systems rather than concentrating solely on offenders after victims have transferred money. The measures were agreed through the Five Country Ministerial involving Australia, Canada, New Zealand, the United Kingdom, and the United States.

The countries have backed stronger action against cross-border fraud and online scams, including increased support for the INTERPOL Global Fraud Taskforce. The aim is to map organised networks, connect criminal groups operating across jurisdictions, develop intelligence, and coordinate disruption rather than leaving national police forces to investigate individual cases whose infrastructure and organisers may sit thousands of miles away.

The communiqué places particular attention on overseas scam centres, which authorities connect with illicit finance, trafficking, and other organised crime. Governments intend to share more information, work with countries hosting those operations, and deepen practical relationships with businesses controlling the communications, platform, and financial infrastructure used to reach victims and move funds.

Artificial intelligence is adding another layer to that model. Ministers identify AI-enabled technology among the capabilities supporting changing fraud operations, giving criminal groups cheaper tools for generating convincing content while they continue relying on familiar telecoms, platform, identity, and payment infrastructure.

Fraud prevention is moving upstream

Modern scams rarely depend on a single provider. A campaign may begin with an advertisement or social-media account, move to a messaging service or telephone network, use stolen personal information to establish credibility, and finish through a bank transfer, payment service, cryptocurrency exchange, or network of intermediary accounts.

No organisation sees that complete sequence on its own. A platform can detect a fraudulent advertisement without knowing whether a resulting bank transfer succeeds, while a financial institution may identify suspicious payments without seeing the account, number, or messaging infrastructure used to manipulate the customer.

Britain’s 2026–29 Fraud Strategy already adopts a system-wide approach intended to close some of those gaps. More than £250 million has been committed over three years, including an Online Crime Centre bringing together expertise from government, policing, intelligence agencies, banks, mobile operators, and technology companies.

The Five Eyes agreement extends that logic internationally. Well-integrated domestic systems remain limited when scam compounds operate abroad, online services are headquartered in another jurisdiction, and stolen funds move through several countries before investigators identify the network behind them.

International cooperation can make infrastructure disruption more practical if authorities can identify the same criminal entities across payment, telecoms, and platform datasets. It also creates legal and operational questions around how quickly communications records, customer information, and financial intelligence can be shared across jurisdictions with different safeguards.

AI adds volume to established methods

Generative AI can help fraud groups produce persuasive messages, translate scripts, clone voices, generate synthetic imagery, and adjust approaches more cheaply, but many of the systems needed to turn manipulation into money remain conventional. Phone numbers, domain names, bank accounts, payment routes, advertising systems, and money-mule networks still create points at which providers and investigators can intervene.

That makes infrastructure cooperation potentially more consequential than attempting to detect every AI-generated message. Synthetic-content detection becomes harder as models improve, whereas disabling a known fraudulent account, blocking a payment route, or disrupting a scam-centre network can impose costs on the underlying operation regardless of which model generated the latest script.

The agreement does not create a common statutory regime forcing platforms, telecoms companies, and banks to use identical controls. Its emphasis is on operational cooperation, intelligence exchange, industry partnership, and international disruption, so the results will depend on whether participating organisations can turn shared information into sufficiently rapid action across different legal and commercial systems.

The wider ministerial meeting also links AI with national-security questions, including cooperation with industry and access to advanced models for cyber-security purposes. Fraud offers a more immediate operational test because the companies involved already control systems through which criminal activity travels.

Telecoms providers can identify suspicious calling behaviour, banks can interrupt payments, platforms can remove accounts and advertisements, and technology companies can investigate abuse of their services. Those interventions become more useful when signals can be connected before criminals simply move to another provider.

Expanded INTERPOL work is intended to add another coordination layer by mapping networks rather than treating each approach to a victim as an isolated offence. Where several scams can be connected to common infrastructure or operators, law enforcement can concentrate on shared dependencies rather than repeatedly dealing with disposable accounts.

The Five Eyes countries now have a broader framework for that cooperation, while Britain already has a domestic mechanism intended to put companies and agencies into a shared operational process. Its value will be measured by whether intelligence moves quickly enough between platforms, telecoms networks, banks, governments, and overseas law enforcement to interrupt fraud while the infrastructure remains active.