Artificial intelligence (AI) is increasingly being used by scammers to deceive people. The warning comes from Prasad Calyam, a professor of engineering and director of the AI research center at the University of Missouri (USA). According to him, spotting deception is becoming harder: whereas scammers used to be given away by spelling errors and clumsy text, neural networks now write smoothly and convincingly.
With the help of AI, criminals create convincing phishing emails (messages that extract data under the guise of official ones), fake online stores, and also clone the voices of acquaintances and record deepfake videos (fake clips in which a person looks and speaks like the real one). Another threat is chatbots that pose as support service employees or romantic partners. They hold a natural conversation, answer questions and gradually gain trust, and then ask for money or confidential information. Because AI adapts to the interlocutor, such conversations are harder to recognize as fraud.
Calyam advises not sharing sensitive information with AI tools: social security numbers, home address, passwords, bank details, medical records and confidential work documents. Seemingly harmless data is also dangerous - date of birth, travel plans, information about relatives. Depending on the platform, information may be stored, viewed or used to train the model. If it leaks, criminals may use it for identity theft, account hacking or personal deception. Before uploading documents, Calyam recommends removing sensitive details and checking privacy settings.
How can you tell that an email, message or call is fraudulent if AI makes them increasingly convincing? Traditional signs such as typos and poor language are becoming unreliable. Instead, you should be wary of unexpected requests for money, personal data or confirmation codes, especially if the message creates a sense of urgency or demands secrecy. You should not follow unfamiliar links or reply to suspicious messages. If something seems doubtful, it is better to contact the organization through its official website or call the person on a verified number.
To protect personal data, the expert advises using a password manager (a program that creates and stores complex passwords) and enabling multi-factor authentication (confirming login through two or more independent methods) on important accounts. Passkeys (digital access keys instead of a password) and hardware security keys provide additional protection against phishing. It is worth checking the privacy settings of AI platforms and disabling optional data sharing for model training if such an option exists. It is also important to update devices and applications, not follow suspicious links and monitor unusual activity on financial accounts.
If there is a suspicion that an account has been hacked, the first thing to do is change the password through the official website or app and log out on unfamiliar devices. If the same password was used in other services, it must be changed there too. Pay special attention to your main email: through it, attackers can reset passwords for other services. If financial data leaks, you should contact the bank or the credit card issuer. If identity theft is suspected, Calyam advises visiting the IdentityTheft.gov website, where recommendations for recovery can be obtained.