How to Spot an AI Dating Bot: 5 Romance Scam Signs

Techwalla may earn compensation through affiliate links in this story. Learn more about our affiliate and product review process here.

How to Spot an AI Dating Bot: 5 Romance Scam Signs

If you're trying to figure out how to spot an AI dating bot, you're probably asking the wrong question. The FBI and FTC's latest guidance describes something different: a human scammer using AI tools to build a fake identity that's harder to catch than it used to be. The FTC warns that AI and other technology mean a romance scammer "might not be so easy to spot" (FTC, earlier this year). The FBI's Birmingham and San Antonio field offices back that up: criminals are using generative AI to smooth out translations, generate realistic profile photos, and produce short video clips meant to prove a contact is real (FBI Birmingham, earlier this year).

First, separate two ideas. People use "AI dating bot" to describe several different things: automated software built to chat without a person behind it, or a human scammer using AI-generated material to build a fake identity. The guidance cited throughout this piece concerns the second problem, fake identities built to manipulate people into sending money or sharing information, including cases in which a scammer leans on AI tools to make that identity more convincing (FBI Birmingham, earlier this year). This guide uses "AI-assisted romance scam" for the threat that guidance documents.

What follows are dating-app red flags that matter whether the account is automated, operated by one person, or managed by a criminal group: five warning signs, in the order they typically unfold, plus how to verify who's actually there and what to do once something feels wrong. If you're looking for signs you're talking to an AI bot, keep in mind that polished language and clean photos are weak evidence on their own. None of this depends on catching a glitch. It depends on recognizing a pattern before money, access, or trust changes hands.

Advertisement

Why polished photos and messages are no longer reliable proof

Video of the Day

The FBI's San Antonio field office puts it plainly: AI-generated content has gotten sophisticated enough that it's "increasingly difficult to distinguish fabricated materials from legitimate information" (FBI San Antonio, earlier this year). Traditional advice on spotting a scam often focused on bad grammar or a stiff, generic profile photo. That advice alone doesn't cut it anymore.

The FBI's Birmingham office has documented the specific ways criminals now use AI in these schemes: smoothing out translations so language mistakes don't give them away, generating realistic images for fake social media profiles, sharing AI-made photos privately to build a sense of intimacy, and producing short videos sent in private messages to "prove" the person on the other end is real (FBI Birmingham, earlier this year). Notice what that last one isn't. It's not a live, two-way video call. It's a clip, sent on the scammer's schedule, which is a different kind of proof than real-time contact.

The stakes are already climbing. Reported losses in the FBI's San Antonio Division jumped from $15.8 million in 2024 to more than $28 million in 2025, and Alabama logged 261 reported cases totaling $7.56 million in 2025, though many losses go unreported (FBI San Antonio, earlier this year; FBI Birmingham, earlier this year). Those are regional numbers, not a national total.

What hasn't changed is the underlying scheme. Someone builds a fake identity to earn trust, then uses that trust to get money, personal information, or account access (FBI Birmingham, earlier this year). AI changes the packaging, not the goal, which is why behavior matters more than any photo-checking trick.

Advertisement

Video of the Day

How to spot an AI dating bot by watching the pattern

Here are the AI romance scam warning signs that matter, laid out as a progression rather than a checklist. A single sign might have an innocent explanation. Several appearing together are more concerning than any one sign alone.

1. Rapid, intense intimacy. Scammers profess love quickly and push the relationship forward faster than the timeline would suggest, often coming across as unusually perfect or emotionally intense from the first few messages (FBI San Antonio, earlier this year; FBI Birmingham, earlier this year). Slow the pace on purpose, avoid commitments, and see whether the person respects your boundaries.

2. Pressure to go off-platform and cut out other opinions. Moving off a dating platform isn't proof of anything by itself, but scammers frequently push for it specifically to avoid detection, along with demands for exclusivity, secrecy about the relationship or any money that's come up, and discouragement from talking to friends or family (FBI San Antonio, earlier this year; FBI Birmingham, earlier this year). Isolation can remove the outside perspective that might otherwise catch inconsistencies (FBI San Antonio, earlier this year). Staying on the platform a while longer is a precaution, not a guarantee; tell at least one person outside the relationship what's going on regardless.

3. Avoidance of verifiable, real-world contact. Common excuses involve military deployment, an oil-rig job, or overseas work, paired with meetups that get planned and then repeatedly cancelled (FTC, earlier this year). Avoiding live video chat is a related flag the FBI calls out directly (FBI Birmingham, earlier this year). A single missed meetup isn't proof of anything; repeated cancellations, especially alongside secrecy, evasiveness, or a money request, are more concerning. An emergency story may be used to set up a money request, so don't send money to resolve one, whatever the explanation.

4. Inconsistent stories and escalating crises. Watch for details that don't line up, or that sound too good to be true, along with a string of dramatic emergencies, a death in the family, sudden medical bills, a failed business, all used to keep emotional pressure on (FBI Birmingham, earlier this year; FBI San Antonio, earlier this year). If details are hard to track, write down names, dates, locations, and jobs as they come up, so you can compare them later.

5. A request for money, access, or an "investment." The request may not surface for weeks or months, but it almost always comes (FBI San Antonio, earlier this year), usually through gift cards, wire transfers, prepaid cards, or cryptocurrency, because those are hard to claw back once sent (FTC, earlier this year; FBI Birmingham, earlier this year). The FBI's San Antonio office expects continued growth in international cryptocurrency investment scams, sometimes called "pig butchering," where a romantic contact steers a victim toward fake high-return crypto investments that can never be withdrawn (FBI San Antonio, earlier this year). The ask isn't always cash, either: some victims are pressed to hand over financial account access or unknowingly move money as a mule (FBI San Antonio, earlier this year). Stop the transaction the moment it's requested. The FBI advises never sending money, gifts, cryptocurrency, investments, or valuables to anyone met only online, regardless of the amount or the explanation attached to it (FBI Birmingham, earlier this year).

No single stage proves a scam on its own. It's the stages together, not any single behavior, that describe the arc a scammer moves through, whether or not AI is doing the polishing.

Advertisement

Advertisement

How to tell if an online dating profile is fake

A few checks can help confirm who's actually behind the account, though none of them settle the question alone.

Run a reverse image search on the profile photo. Scammers frequently recycle stolen images across multiple fake profiles, and a search can surface another name or details that don't match (FTC, earlier this year; FBI San Antonio, earlier this year).

Search their stated occupation along with the word "scammer." It's a small step, and it may turn up similar reports from other people who ran into the same story (FTC, earlier this year).

Verify identity and claims through sources that are independent of anything the person has sent you directly (FBI San Antonio, earlier this year). Don't treat a link, screenshot, or investment platform the person recommends as independent verification; it comes from the same source you're trying to check.

Consistent, independently verifiable details are a useful positive signal, but they don't authenticate anyone by themselves. Treat each check as one data point, not a final answer. A photo that reverse-searches clean, or a video clip meant to prove identity, doesn't clear someone; those are two facts among several. Stack multiple independent checks, image search, occupation search, consistency of details over time, before trusting the relationship with money, account access, or anything sensitive.

Advertisement

Advertisement

When to act, and what to do

Set a threshold before you need one. A single canceled date has plenty of innocent explanations. Repeated cancellations paired with secrecy, emotional pressure, or a request for money is far more concerning, and that's the point to stop and reassess.

Once that threshold is crossed, stop contact immediately (FBI San Antonio, earlier this year), and talk to someone you trust to see if they share the concern (FTC, earlier this year). Don't delete the messages. The FBI advises keeping every communication, since it's evidence if the case goes anywhere (FBI San Antonio, earlier this year).

If the scammer talked you into installing software, disconnect from Wi-Fi or the internet first, then shut the computer down before reporting the incident (FBI San Antonio, earlier this year).

If money has already changed hands, contact the financial institution right away (FBI San Antonio, earlier this year). Report the incident to the dating platform itself, to the FTC at ReportFraud.ftc.gov (FTC, in 2024), and to the FBI's Internet Crime Complaint Center at IC3.gov (FBI Birmingham, earlier this year).

Report it sooner rather than later. As Birmingham's field office puts it, prompt reporting helps investigators track criminal networks, hold offenders accountable, and keep others from becoming victims (FBI Birmingham, earlier this year).

Advertisement

Advertisement

What this means going forward

AI hasn't invented a new scam. It's given an old one sharper tools: cleaner grammar, more convincing photos, a video clip meant to make the contact look real (FBI Birmingham, earlier this year). What those tools disguise hasn't changed: speed, isolation, and an eventual ask.

That's why a behavioral checklist ages better than a visual one. A chatbot or a stolen photo can't fake how fast someone says "I love you," whether they'll get on a live, unscripted call instead of sending a clip, or what happens the moment money comes up.

There's no need to prove that a photo, message, or video was made with AI. If someone is rushing intimacy, isolating you, avoiding ordinary verification, and asking for money or access, that pattern is the evidence. Stop contact, preserve the messages, and report the account.

Advertisement

Advertisement