AI Dating Scams: How Fake AI Profiles Work and How to Spot Them
Fake dating profiles are not new, but AI has made them cheaper to run and harder to notice. Here is what has been reported, and what to watch for.
What has been reported
Anthropic's September 2026 threat intelligence report describes a network of fake dating apps built to defraud users. News coverage of the report says the apps ran more than 20 dating apps, that roughly 75% of matches were AI-generated personas, and that more than 4,700 personas talked with at least 25,000 people over two weeks in April 2026, exchanging around 2.36 million messages. According to that coverage, users were not told they were talking to machines.
The same reporting says matching and messaging used up a limited quota that users refilled by buying virtual coins, and that paid workers appeared on video calls to make the profiles look real. These figures come from news coverage of the report, so treat them as reported, not as our own findings.
It isn't a new tactic
In 2016 the US Federal Trade Commission and 13 states settled with the operators of Ashley Madison. The FTC charged that the site used fake "engager" profiles, so people paid to message what they believed were real women. The order carried an $8.75 million judgment, partly suspended. What changed is the tool: AI makes each fake profile far cheaper to run.
Warning signs
- Replies that arrive instantly at any hour and feel warm but generic.
- A match who keeps finding reasons not to video chat.
- Pressure to spend at emotional moments, such as a flood of likes or profile views right as your free messages run out.
- Details that shift or contradict earlier answers.
- A profile that vanishes or goes silent once you stop paying.
A video call is not proof
According to the reporting, paid workers appeared on video calls while another AI model suggested their replies. Seeing someone on camera shows a person exists. It does not show that person is writing the messages. Photo verification has the same limit: it can confirm a photo matches a face, not who is typing.
Charging isn't the problem. Hiding is.
Paying to send more messages is not automatically a scam. The question is disclosure: are you told who you are talking to, and what exactly you're paying for? If an app hides that, stop spending.
What to do if you think you've been targeted
- Stop buying coins, tokens or credits.
- Screenshot the profile and the conversation.
- Report the app to the app store and to the platform.
- Report the fraud to your consumer protection authority. In the US, that's reportfraud.ftc.gov.
- Ask your bank or card provider about a dispute for recent purchases.
How 1n2mate approaches this
- Profiles on 1n2mate are created by members. We don't create fake profiles or use AI to pose as members.
- You start with 15 free messages to send. Extra credits are $5 for 40, they don't expire, and there is no subscription.
- Chat blocks phone numbers, emails, links and handles, and every profile has one-tap Block and Report, including a "Bot or scam" reason.
- Anyone from the 1n2mate team who talks with you is identified as team, never as a member.