Analyze the situation. Return JSON {"scam_signs":[...],"safe_action":"...","verify_channel":"..."}: scam_signs - signs of fraud; safe_action — what to do instead of paying; verify_channel - how exactly to check the request. Return only JSON, no text around it.
Mechanics: AI has made mass deception cheaper - now letters are error-free, in your language, with details from social networks, and the voice of the “boss” or “relative” is cloned in a couple of minutes of audio. The scheme is always the same: to evoke emotion (fear, greed, urgency) and lead away from verification. Signs: unexpected urgency, request for money, codes or transfer, new or slightly changed sender address, “log in and confirm” link, secrecy (“don’t tell anyone”). “Stop channel” reception: check any monetary or “urgent” request through ANOTHER channel - a letter has arrived, call; If you called, write to a known contact. The fraudster only controls the channel where he started. The second trick: never follow a link from an email - go to the site manually, through a bookmark or search. Insider: codes from SMS are a password, they are not dictated to anyone, not even to the “bank security service”; a real bank never asks for a code. A typical mistake: thinking “I won’t be deceived” - the victims are more often the confident ones, not the naive ones.
Unlock access to submit solutions for instant AI review.