Scammers have always relied on creating a sense of trust, urgency or panic. However artificial intelligence is now giving them a powerful new tool and is making scams harder to spot.
AI can now produce convincing emails, messages and social media posts, mimic the tone of a genuine company and create remarkably plausible conversations. That means some of the traditional clues we have relied on to spot a scam such as poor spelling, awkward wording or suspicious messages are becoming much less reliable.
As a Consumer Champion, I’m particularly concerned about what this means when people are trying to resolve a problem with a company. Someone who has posted publicly about a missing delivery, a refund or a complaint may suddenly be contacted by an account claiming it can help. In an AI-driven world, that conversation could look incredibly genuine.
So how can you protect yourself? Here are 10 ways to spot and avoid AI scams.
Start from the company’s official website
This is one of the simplest and most effective ways to protect yourself. If you need to contact a company, don’t just click on a customer service link sent to you through social media, WhatsApp, email or a message from someone claiming to work for the company. Instead, type the company’s website address yourself or use an app you already have. From there you can lcoate the official customer service or complaints details.
Scammers can create remarkably convincing customer service profiles and links so don’t assume that a professional-looking page means it is genuine.
Be suspicious of unsolicited ‘help’
Have you ever posted a complaint on social media and suddenly received a message from someone offering to sort it out? Stop before you respond.
Genuine companies may sometimes contact customers through social media, but scammers know that someone who is already frustrated is likely to be particularly receptive to an offer of help. That public complaint can effectively become an invitation to a scammer.
If someone approaches you unexpectedly, don’t use the contact details they provide. Go back to the company’s official website and contact the company independently.
Don’t be misled by the name, logo or profile picture
A company logo doesn’t prove that an account belongs to the company. Scammers can copy names, branding, profile photographs and other details from genuine businesses. AI makes it easier to create convincing content around those fake identities too.
Look beyond the branding. Check the account handle, how long the account has existed, its previous posts and whether it links back to the company’s genuine website. Always bear in mind that a blue tick or other verification-style imagery shouldn’t automatically convince you that you’re dealing with the right person.
Don’t move the conversation
One of the biggest red flags is being asked to move the conversation somewhere else. You might be told: “DM me,” “message us on WhatsApp,” “contact this number” or “we’ll sort it out privately.” Don’t automatically follow those instructions.
Before moving from one platform to another, independently find the company’s genuine contact details and use those instead. If you started your complaint through the company’s official website, there’s usually no reason to abandon that verified route simply because somebody has appeared offering an alternative.
Find out whether you’re talking to a human or AI
There’s nothing wrong with asking whether you are speaking to a bot or an agent. If you’re dealing with an online customer service system, ask: You can then decide whether you’re happy to continue or whether you want to speak to a person.
Don’t assume that because the conversation feels natural, there must be a human sitting at the other end. AI can produce remarkably fluent and context-aware responses. A conversation can feel personal without necessarily being personal.
Don’t mistake a convincing conversation for proof it’s genuine
This is one of the biggest changes AI has brought to scams. We have traditionally been told to look for bad spelling, strange grammar or messages that don’t sound quite right. However AI can produce polished, professional and persuasive communications in seconds.
A scammer can use AI to write an email that sounds as though it has come from a bank, airline, retailer or delivery company. They can also use AI to maintain a conversation, respond to questions and make their messages appear increasingly convincing. Good grammar is no longer proof that someone is genuine. Neither are empathy, professionalism or a detailed response. Don’t judge authenticity by how human the conversation feels.
Never share security information
A genuine company may need to verify your identity, but there is a big difference between legitimate security checks and handing over sensitive information. Be extremely cautious if somebody asks for your password, PIN, one-time passcode, full card details or banking credentials.
Never read out a verification code that has just arrived on your phone because someone claiming to be customer services asks you to. If you’re unsure why a company needs particular information, stop and contact it through an independently verified channel.
Verify information independently
This is one of the best habits you can develop. If someone gives you a telephone number, email address, payment link or website, don’t automatically use it. Instead, find the company’s contact details yourself and ask whether the person, case number or proposed procedure is genuine.
In other words, use a different route to verify the first route. It might take a few extra minutes, but those minutes could save you from losing money or handing over valuable personal information.
Don’t pay to get a refund or complaint ‘unblocked’
This is a particularly nasty scam because it targets people who are already expecting money. You may be told that your refund, compensation payment or account recovery requires a fee, deposit or ‘verification payment’.
Don’t make the payment simply because someone says you need to do so. Go back to the company’s official website and check its refund or complaints process yourself. If you’re expecting compensation or a refund, independently contact the company and ask what is happening. You should never feel pressured into paying money simply because somebody has appeared claiming they can release money that is already owed to you.
If you’re not sure, stop and think
You don’t have to keep responding just because somebody claims to be customer services. If something doesn’t feel right, stop the conversation. Tell them you’ll contact the company through its official channel instead and then do exactly that.
A genuine customer service team shouldn’t object to you independently verifying who you’re dealing with. Don’t worry about appearing rude or difficult. Being irritating or inconvenient is much better than being scammed.
Final thoughts
In a nutshell, the biggest change AI is bringing to scams is that we can no longer rely on our instinct that something ‘looks like a scam’. A message can be perfectly written. A profile can look professional. A conversation can feel warm, helpful and completely authentic. That doesn’t mean it is, never trust the conversation, always verify the contact.
If you independently access the company’s official website, use its published contact details and initiate the conversation yourself, you’re starting from a much safer position than if somebody suddenly appears in your replies and invites you into a private conversation.
And if you’re making a complaint publicly, remember that scammers may be watching for exactly that opportunity. A post saying “I’ve been trying to get a refund from X for three weeks” could effectively advertise that you’re a potential target.
In the age of AI, don’t judge whether something is genuine by how human it sounds. Check who you’re actually dealing with. It can be a very good tool when used effectively and responsibly but in the wrong hands it can case no end of very costly issues.
