Fraudsters are increasingly using familiar digital services to deceive people. A fake building residents’ chat, a convincing social media profile, or a recommendation from an AI system can lead to the loss of money and personal data. Artificial intelligence helps criminals create believable messages, imitate voices, and promote fraudulent websites. At the same time, the main tool of deception remains the same: a person’s trust in what looks familiar and reliable.

Fake Building Chats: Scammers Posing as Neighbors

Scammers create fake resident chats or join real groups, especially in new residential complexes where neighbors do not yet know each other well. The reason for reaching out usually looks quite ordinary: voting on neighborhood improvements, registering in the building chat, verifying data with the property management company, or receiving a new intercom key.

Participants are then asked to follow a link, scan a QR code, install an app, or share a code from an SMS message. Under the pretext of identity verification, scammers try to gain access to the victim’s account. Depending on the situation, this could be an account in a messenger app or another service.

Once they gain control of the account, criminals can message the owner’s contacts on their behalf, distribute malicious links, and ask for money. This is how the attack spreads further: messages from a familiar person raise less suspicion than messages from an unknown number.

Factories of Trust: AI Helps Create Convincing Identities

Scammers have also become more sophisticated in using fake profiles designed to look like pages of real people. These may include photos, posts, comments, reviews, and a history of activity. Such accounts are created to gradually win the trust of a potential victim.

Generative artificial intelligence makes this work easier. It can be used to quickly prepare believable texts, images, avatars, and voice messages. Criminals find it easier to maintain conversations, tailor messages to a specific person, and run several fraudulent campaigns at the same time.

The F-Secure Scam Intelligence & Impacts Report 2026 states that 89% of cases in which scammers use AI are related to improving the “bait” — messages, images, audio, and video intended to convince a potential victim. This is a figure from F-Secure’s research, not the share of all fraudulent operations in which artificial intelligence is used.

Fake voices pose a separate threat. With the help of AI, criminals can imitate the voice of a relative, colleague, or another familiar person in order to obtain an urgent money transfer or confidential information. In February 2026, the UK’s National Trading Standards reported a fraud scheme in which criminals used voice cloning to set up unauthorized recurring withdrawals from bank accounts.

When AI Recommends Scammers: Fake Websites in AI Responses

Another, perhaps the most relevant emerging threat is linked to the fact that people are increasingly turning to AI assistants for advice, company contacts, and product recommendations. Criminals are trying to exploit this trust by getting AI systems to link to fake websites and provide false information.

To do this, they may use copies of websites belonging to well-known companies, addresses resembling the original ones, fake reviews, and specially prepared materials. If a user finds such a resource through an AI response, they may take the recommendation as confirmation of the seller’s reliability.

In June 2026, the UK’s National Trading Standards warned that fake clone websites had appeared in AI search responses and were receiving hundreds of visitors a day. Such websites imitated real stores and accepted payment for goods that buyers never later received.

F-Secure also describes the risk of manipulating the data that AI systems rely on when generating responses. In this field, the term AI data poisoning is used. Depending on the specific method, this may involve tampering with the data used by the model or attempts to influence the information it finds and uses when answering.