Investigators say U.S.-built artificial intelligence and internet backbones are helping overseas scam compounds steal billions from Americans, and much of it runs in plain sight.
Story Snapshot
- American-made AI tools and U.S. network providers appear inside global scam operations.
- One in five device signals from Myanmar scam compounds rode U.S.-registered networks.
- The Treasury Department sanctioned Southeast Asian networks tied to romance fraud.
- Government analysts say Americans lost at least $10 billion to these scams in 2024.
What the new investigation uncovered about U.S. tech in scams
Associated Press reporters, working with a Washington security nonprofit, found scammers using American-made artificial intelligence models to build software that handles many languages and tracks worker output. The team also mapped how devices inside Myanmar scam compounds connected to the wider web. The analysis tied traffic to U.S.-registered internet providers at notable rates. These findings show how tools built for good can be bent to do harm by organized crime abroad.
The same reporting listed several U.S. companies whose services helped carry or host the data that powers these schemes. Names included major network carriers and hosting firms. The outlet said the evidence came from tens of thousands of leaked files and more than 200,000 device connections gathered by an anti-trafficking group. The picture is not one platform or one app, but a supply chain that stretches from cloud tools to cables and satellites.
Scale of losses and how the scams work on victims
A federal commission that tracks risks tied to China-linked activities reported at least $10 billion in American losses in 2024 from Southeast Asia-based scams. Analysts there warned that losses could grow in 2025 if nothing changes. The same materials describe a spin on “pig butchering” scams. Criminals use translation and chat tools to pose as romantic partners. They groom trust. Then they push fake investments or sudden crises to draw out savings.
The United States Department of the Treasury announced sanctions on dozens of people and firms across Southeast Asia tied to online romance and investment fraud. The move aims to cut these actors off from the American financial system and warn banks worldwide. Sanctions can freeze assets and block transactions. They also send a clear signal that the government sees these networks as global and dangerous. But sanctions alone do not stop the internet routes at the heart of the trade.
Why the infrastructure story matters to both left and right
Many Americans, across party lines, feel big institutions protect themselves first. This story feeds that worry. The investigation shows how foreign crime groups leaned on American innovation and bandwidth to reach U.S. phones. Companies designed tools to help people work and talk, not to cheat seniors out of savings. Yet the system did not prevent that misuse. The result is a sense that rules lag far behind criminal tactics that move fast online.
Truth Desk Media Report 1: Hybrid Scams Surge in Australia – $2.18 Billion Lost as Transnational Networks Expand Reach
⁰July 5, 2026
Australians lost a staggering $2.18 billion to scams in 2025, marking a 7.8% increase from the prior year despite some reporting dips. Hybrid…— Truth Desk (@Nizibizizbig) July 5, 2026
Conservatives see a failure to police borders in cyberspace. Liberals see corporations profiting while ordinary people get hurt. Both see a government response that looks split across agencies. A bipartisan commission called current efforts “fragmented” and “under-resourced.” That critique suggests no single office owns the fix. It also hints at a deeper problem: criminals use the same global networks that power everyday life, which makes blunt crackdowns risky and slow.
Gaps in proof and how to read the claims carefully
The reporters link American-made artificial intelligence to scam software by showing features and outputs. They did not share raw logs that name specific accounts at model providers. That means we know how the tools likely worked, but we do not see a binding trail from a given scammer to a given account. Still, the network routing data and the sanctions records provide strong, on-the-record anchors for the broader claims about U.S. tech inside the pipeline.
Readers should also separate two ideas. First, the infrastructure and tools are often American. Second, the groups running the compounds are based in Southeast Asia with reported links to Chinese criminal syndicates. The first point rests on technical findings and sanctions actions. The second often comes through policy reports and briefings. Both can be true, but they draw on different kinds of evidence and carry different levels of legal certainty.
What practical steps could protect people now
Lawmakers could push for speed limits on risky features, like mass translation and persona generation, when used at scale. Companies could add stronger checks on automated outreach and bulk account creation. Network firms could improve detection of traffic from known scam hubs and flag patterns to banks and platforms. These are narrow, focused moves that do not break the open internet, but they make it harder to run a sweatshop scam farm on top of U.S. tech.
Families can take simple steps. Verify voices and video with a live, unique action on camera. Refuse investment advice that starts in a romance chat. Pause when someone you just met urges secrecy. Use bank-level safeguards and two-factor logins. Report losses fast to your bank and local police. These small defenses matter when criminals use advanced tools. They buy time. They also create more signals that help investigators connect cases across borders.
Sources:
pandasecurity.com, abcnews.com, linkedin.com
