AI fraud is the use of artificial intelligence to make scams more realistic and harder to recognize. Scammers can use AI to imitate voices, create convincing images and videos, write polished messages or build fake websites — all designed to gain your trust and potentially steal money or personal information.
The holiday season often brings the moments we look forward to all year: finding the perfect gift, gathering with family and friends and checking items off our shopping lists. Unfortunately, it also brings an increase in fraud and scam activity.
Many of the tactics themselves aren't new. What's changing are the tools fraudsters can use to make them more believable. A message can look professional. A website can appear legitimate. Even a familiar voice may not be what it seems.
According to the FBI's Internet Crime Complaint Center (IC3), consumers and businesses across the United States reported more than $20 billion in internet crime losses in 2025, highlighting the growing scale and sophistication of fraud nationwide.
The technology may be changing, but the objective remains the same: gain your trust, create urgency and persuade you to act before taking time to think. That makes independent verification increasingly important, even when a request looks or sounds real.
Traditional scams often contained clues that something wasn't right: awkward language, spelling mistakes, unusual formatting or a message that simply didn't sound like the person or company supposedly sending it.
Artificial intelligence can make some of those clues less obvious.
Generative AI can help fraudsters write polished emails and text messages, while voice-cloning technology can make a caller sound like someone you know. Deepfake technology can create realistic-looking video or audio, and AI can also help criminals develop professional-looking websites, advertisements and online profiles.
These tools can support social engineering, in which scammers manipulate people into sharing information, sending money or taking another action. The result is an important shift in how we think about fraud prevention: something looking professional or sounding familiar isn't necessarily proof that it's legitimate.
AI-powered scams can take many forms, but the goal is often similar: create enough trust or urgency to convince the victim to act before verifying the request.
Here are some scenarios to watch for.
Let's say you're a grandparent. The phone rings, and your grandchild is on the other end.
They sound upset and say they were in an accident while traveling. They need help and ask if you can send money immediately. They plead with you not to tell anyone else. The voice sounds familiar, the story feels believable and your first instinct is to help.
While scams like these are not new, artificial intelligence is making them more convincing. Fraudsters can use AI voice-cloning technology to imitate a person's voice and create calls designed to appear genuine and earn your trust.
The urgency and secrecy are important warning signs. Before sending money or sharing sensitive information, contact the person directly using a trusted phone number or reach out to another family member to verify the story.
You find the perfect gift online at an incredible price. The website looks professional, the photos are convincing and the customer reviews appear genuine. Excited to secure the deal, you place the order.
Artificial intelligence can help fraudsters build fake websites, realistic product descriptions, convincing advertisements and believable customer reviews. Scammers may also use QR codes in advertisements, emails, social media posts or promotional materials to direct shoppers to fraudulent websites designed to collect payment or personal information.
Before purchasing from a link you found on social media, take a few moments to research the seller and verify the website URL. Exercise caution when scanning QR codes that direct you to payment or login pages.
If a deal seems too good to be true, it probably is.
A package delivery notification arrives by text message.
The message says a package is delayed and updated payment information is needed before delivery can be completed. It includes a convenient link to resolve the issue.
Because many of us are expecting deliveries during the holidays, these messages can appear legitimate. Phishing is a type of scam in which criminals impersonate a trusted person or organization to persuade someone to provide information, click a malicious link or take another action.
Generative AI can make phishing emails and text messages more polished and personalized, potentially making an AI scam more difficult to recognize based on language alone.
When tracking a package, navigate directly to the retailer's or shipping carrier's website rather than clicking links in unexpected messages.
A familiar face can create the same sense of trust as a familiar voice, and deepfake technology can be used to create realistic-looking fake video or audio of another person. In deepfake scams, fraudsters may impersonate public figures, financial professionals, company executives or other trusted individuals.
For example, fabricated videos, endorsements or online profiles could be used to make an investment opportunity appear legitimate. Scammers may combine these materials with fake investment platforms or websites to reinforce the story and encourage a person to transfer money or provide financial information.
Before acting on an investment offer or unexpected financial request, independently verify both the person making the request and the organization involved. A convincing video, voice or online profile should not substitute for verification.
An employee receives an email that appears to come from a trusted supplier. The request seems straightforward: update payment instructions to send a payment to a new account.
Everything looks legitimate.
Artificial intelligence is helping fraudsters create polished business communications that are free of obvious spelling or grammar mistakes. This can make fraudulent payment requests harder to distinguish from legitimate ones.
Before processing payment changes, businesses should independently verify requests using established contact information and follow existing approval procedures.
Because AI-enabled scams can appear polished and realistic, identifying fraud may require looking beyond spelling mistakes or poor-quality images.
Potential warning signs include:
No single warning sign necessarily means a communication is fraudulent. At the same time, the absence of obvious mistakes doesn't establish that a request is legitimate. When something unexpected involves money, account information or personal data, taking time to verify it through a separate, trusted channel can help you determine what you're really dealing with.
As AI technology evolves, the fundamentals of fraud prevention remain useful: pause, verify and protect your information.
Here are five practical steps you can take.
If you receive an unexpected call, email or text asking you to send money, change payment information or provide account details, independently verify the request.
Contact the person or organization using a phone number or website you already know or have independently located. Don't rely on the phone number, link or contact details provided in the suspicious message itself.
This extra step can be particularly important when a request involves an urgent money transfer or changes to banking information.
Consider establishing a family code phrase that can be used to verify someone's identity during an emergency request.
If a caller claims a relative is in trouble, contact that person directly or check with another family member before taking action. With voice cloning making it possible to imitate familiar voices, hearing someone who sounds like a loved one may no longer be enough to confirm their identity.
Never share passwords, one-time passcodes or verification codes with someone who contacts you unexpectedly.
Be cautious about entering account or financial information through unfamiliar links, and use available account security features. Regularly reviewing your account activity may also help you identify transactions you don't recognize.
For more practical steps, review Cathay Bank's cybersecurity tips to help you stay safe and information about bank cybersecurity trends and protection.
Instead of following an unexpected link, navigate directly to the organization's official website.
Research unfamiliar retailers and investment opportunities before providing payment information. Check website addresses carefully, particularly before entering login credentials, banking details or other sensitive data. QR codes deserve the same caution as other links. You may not know where a QR code will take you until you scan it.
Think about how much personal information, including photos, videos and audio recordings, is publicly available through social media.
Scammers may use publicly available data to make impersonation attempts more convincing. Reviewing privacy settings and being thoughtful about what you share publicly can reduce the amount of information readily available to someone attempting to build a convincing story about you or someone you know.
If something doesn't feel right, stop before taking the requested action. Don't click additional links, send money or provide more information. Instead, independently contact the person, business or financial institution supposedly making the request using trusted contact information.
If the communication involves your Cathay Bank account, banking information or a transfer, contact Cathay Bank using a trusted number, such as the number on the back of your card.
If you've already shared information or believe fraudulent activity may have occurred, contact your financial institution for guidance on appropriate next steps. You can also learn more about how to protect yourself when banking online.
Scams and other internet crimes can be reported to the FBI's Internet Crime Complaint Center.
AI fraud involves using artificial intelligence to facilitate or make fraudulent activity more convincing. It may include AI-generated messages, voice cloning, deepfakes, fake websites and other techniques intended to deceive people into sharing information, sending money or taking another action.
Yes. Voice-cloning technology can create audio that imitates a person's voice. Scammers may use this technology in impersonation scams, including fake emergency calls involving family members. Independently contacting the person through a trusted number can help you verify whether the request is genuine.
There may not always be an obvious visual or grammatical clue. Watch for behaviors commonly associated with scams, including unusual urgency, secrecy, unexpected requests for money or information and pressure to bypass normal verification.
When you're uncertain, independently contact the person or organization through a trusted channel rather than replying directly to the suspicious communication.
Stop interacting with the suspicious communication and independently verify the request. If the situation involves banking or financial information, contact your financial institution through a trusted number for guidance on appropriate next steps.
Whether it is a familiar voice, an online shopping bargain, a package notification, a deepfake or a business payment request, scammers often succeed by creating urgency and encouraging quick decisions.
Artificial intelligence is making scams more convincing. From realistic voices to sophisticated websites and polished emails, not everything that looks real is real.
If you receive an unexpected call, text message, email or notification, take a moment to verify before taking action. Contact organizations through trusted phone numbers or websites, and never share one-time passcodes or verification codes with someone who contacts you unexpectedly.
The holidays should be about spending time with family and friends, not recovering from fraud.
If something feels off, trust your instincts. Reach out to Cathay Bank by calling the number on the back of your card. When in doubt, call us first.
This article does not constitute legal, accounting or other professional advice. Although the information contained herein is intended to be accurate, Cathay Bank does not assume liability for loss or damage due to reliance on such information.
