ROMANCE SCAM · CATFISHING · AI IMAGES
How to recognise a romance scam: catfishing and AI images
A romance scam builds trust over time before requesting money, information or access. A fake profile may use stolen photos, AI-generated images or video and maintain a plausible story for months.
Combine evidence: requests for money, independent identity checks, image context and conversation. An AI text or image detector result alone cannot prove a person is a scammer.
1. Notice the pattern of emotional manipulation
The person may be attentive, move conversations off the dating platform, repeatedly postpone meeting and eventually cite medical costs, blocked travel or frozen accounts. Each event may have an innocent explanation; together with a money request, they warrant stepping back.
For an employer, the manipulation may target project details, internal access or files. Keep personal and work data separate and offer a discreet, non-judgemental reporting channel.
2. Check messages and shared photos separately
In the widely reported fake Brad Pitt case, Anne described transferring €830,000 to scammers impersonating the actor. TF1 reported fake selfies, forged documents and videos using AI. This illustrates how sustained deception works; victims need support, not ridicule.
To check a message, paste a sufficiently long excerpt without private details into IA Detector. It flags stylistic patterns consistent with AI assistance but cannot identify its author: human, translated or edited writing may look synthetic. AI writing is not automatically fraudulent, and scammers can write their own messages.
For a photo, use SecurCheck image analysis with a non-intimate copy. Metadata, software markers and visual inconsistencies may help, but re-encoding or messaging can remove them. Try reverse image search too: a genuine photo can be stolen or used under a false name. No match does not validate a profile.
Even a video call is not definitive proof of identity: content may be staged or manipulated. Assess the context, requests and details you can verify outside the conversation.
Help executives and teams recognise emotional manipulation and fake profiles
A fake profile may build trust, request money or seek business data and access. Manipulated images and carefully crafted messages can reinforce the deception.
SecurCheck Business helps teams examine messages, stylistic signs of possible AI writing and shared photos separately. Results are clues to combine with behaviour and payment requests, then verify independently.
No AI text or image detector can establish a scammer's identity or prove content origin. Genuine photos can be stolen and synthetic images can evade detection.
Explore the Business offer3. Respond without isolating yourself or exposing access
Stop payments
Do not send more money, buy gift cards or sign crypto transactions under pressure. If you already paid, promptly contact your bank or exchange through usual channels and preserve evidence.
Secure shared information
If you shared logins or work access, change passwords from a clean device, revoke affected sessions and notify your security team. Do not open unexpected files or links from the profile.
Get help and report
Speak to someone you trust, report the profile to its platform and contact the police if you experienced fraud or blackmail. Keep useful evidence without redistributing intimate images. Official services can guide you.
Sources: TF1 Info · Cybermalveillance.gouv.fr.
Conclusion: combine clues without judging
A photo or message that may be AI-generated is one clue among many. Pause transfers, verify identity independently and seek help without shame. Tools help ask better questions, not authenticate a person.