TL;DR
Turn your wind-down time into listening time
- Thousands of audiobooks, podcasts and originals
- Listen on your phone, tablet or Echo — also offline
- Cancel anytime
Anthropic reported that a dating-app operation used Claude to build 4,700 personas that interacted with 25,000 people and sent more than two million messages over two weeks. The company said human gig workers helped with interactions AI could not handle, and the scheme earned money through in-app tokens rather than direct requests for cash.
Anthropic says it found a dating-app operation that used its Claude AI assistant to create 4,700 personas and generate conversations with 25,000 people, sending more than two million messages over two weeks. The case illustrates how AI can be used to make deceptive dating profiles and sustained chats at scale, while the operation’s reported revenue model relied on users buying in-app tokens to continue conversations.
According to Anthropic’s report, criminals worked with a China-based app studio to build a network of 20 dating apps and generate AI personas. Users were led to believe they were communicating with people, though Anthropic said the interactions were not entirely automated. A small number of human gig workers reportedly filled gaps the AI could not, including taking part in brief video calls and following users on social media.
The operation differed from romance scams that persuade victims to transfer money directly to a supposed partner. In this case, the report said, the apps monetized chats through tokens purchased by users. Anthropic counted more than two million messages in two weeks; the source material does not give a total revenue figure or say how much users spent.
The source also cites a McAfee survey in which 35% of Americans who use dating apps said they believed they had encountered fake photos or profiles. That is a survey finding about respondents’ beliefs, not a verified count of fraudulent accounts. Separately, the source reports that U.S. romance scams caused more than $1 billion in losses in 2023, citing authorities, and says a 2025 estimate put total losses from internet crime at nearly $21 billion. Those figures cover different categories and periods.
How AI Changes Dating-App Fraud
The reported network shows how generative AI can reduce the effort needed to maintain many conversations at once. A profile that looks plausible and responds quickly may be harder for users to assess than a static fake account, particularly when occasional human involvement supports the impression that a real person is behind it. The report does not establish that every AI-assisted dating profile is fraudulent, but it describes a scheme designed to mislead users.
The token model also shifts the financial risk. Instead of asking for a large transfer after building a romantic relationship, an app can collect smaller payments as users keep chatting. That can make the commercial incentive less obvious to a user and complicate efforts to distinguish a scam from a paid messaging service. The source does not quantify losses tied to this particular operation, so its financial scale cannot be compared directly with reported romance-scam losses.
For dating platforms, the case raises questions about how they verify profiles, monitor repeated or automated behavior, and explain paid features. The source says several major dating services use AI to detect fake profiles and misconduct, but provides no performance figures or details on how those systems address coordinated networks spanning multiple apps.
As an affiliate, we earn on qualifying purchases.
From Romance Scams to Token Chats
Romance scams have commonly involved fraudsters building trust online before asking victims for money. The source describes a separate prosecution involving six men from Nigeria who were extradited to the United States and allegedly linked to the Black Axe organization. They are accused of using online romances to defraud more than 100 women of over $6 million; those allegations should not be treated as findings of guilt.
The source says authorities recorded more than 64,000 U.S. romance-scam victims and over $1 billion in losses in 2023. It also reports that Visa’s Scam Disruption team identified a network using dating sites and chatbots to draw people into recurring billing. These examples concern different operations. The Anthropic case stands out in the source because it describes AI-generated identities and conversations across a network of apps, with human workers supporting some interactions.
AI has also become part of fraud detection. The source says dating companies including Bumble, Match, Tinder, OkCupid and Plenty of Fish use AI to identify fake profiles or harmful behavior, and that banks and card companies use such systems to spot fraud. It mentions PeopleFinders’ Stud or Dud service as a dating-focused background-check product, but provides no independent evaluation of its accuracy or suitability.
As an affiliate, we earn on qualifying purchases.
What the Report Leaves Open
The supplied account does not identify the dating-app network, say when Anthropic detected or reported it, or specify whether law enforcement has opened a case. It also gives no estimate of user losses, the number of paying users, or whether the apps remain available. The reported counts are attributed to Anthropic, and the source does not provide the underlying report or methodology for independent review.
It is also unclear how many users recognized the conversations as artificial, whether they received refunds, and what happened to the data collected through the apps. The source describes the alleged conduct but does not report charges, convictions or an official response from the app studio or the people involved. Those details should not be inferred from the report’s description.
As an affiliate, we earn on qualifying purchases.
Platform and Law-Enforcement Response
The supplied material does not name a next court date, regulatory action or company announcement connected to the Anthropic report. Further confirmation would depend on additional details from Anthropic, the app operators, dating platforms or law enforcement, including whether the network has been shut down and whether users were notified.
For now, the case puts attention on how platforms detect coordinated accounts and disclose paid messaging systems. The source recommends checking profiles and verifying information, but no single tool or signal can establish that a person is genuine. Users who suspect fraud can report the account through the platform and contact their payment provider about unauthorized or misleading charges.
As an affiliate, we earn on qualifying purchases.
Key Questions
What did Anthropic report?
Anthropic said an operation used Claude to build 4,700 personas that interacted with 25,000 people and generated more than two million messages over two weeks across a network of 20 dating apps.
Did the operation ask users to send money to a romantic partner?
The source says this scheme was monetized through in-app tokens users bought to continue chats. It does not report a total amount collected or say that users were asked to transfer money directly to a supposed partner.
Were the conversations entirely automated?
No. Anthropic’s account, as summarized in the source, says a small number of human gig workers handled interactions such as brief video calls and social media follows that the AI could not perform.
Has anyone been charged over this reported network?
The supplied material does not identify charges, arrests or court proceedings tied to this operation. It describes Anthropic’s findings but does not establish a law-enforcement outcome.
How many dating-app users have encountered fake profiles?
The source cites a McAfee survey in which 35% of U.S. dating-app users surveyed believed they had encountered fake photos or profiles. That measures respondents’ beliefs, not a verified rate of fraudulent accounts.
Source: rss
Fall Picks
fall essentials
As an affiliate, we earn on qualifying purchases.
