A dark side of AI: Text and phone scams are getting worse

AI is changing the way many of us do work, and the same goes for scammers. AI is making their jobs easier —and it’s making us more vulnerable.

Most of us have likely received a scam call or two in the past. (Or even in the past day.) These calls are annoying at best and financially devastating at worst. Some of us may even know someone who has been the victim of a scam call. This year, the average victim of scam calls lost $3,690.

The rapid development of AI technology has made hacking easier because AI tools can scrape our personal data from the internet, store it, and use it in scams. With access to our personal information, AI scams are becoming more effective than ever. 

Here are four ways AI is making scams more difficult to identify — and what you can do to protect yourself. 

How is AI being used to scam people?

#1: More scams all around

Scam calls and texts existed long before AI models came onto the scene, but AI makes them cheaper and more powerful. 

For one, the sheer volume of scam calls is on the rise. The monthly average of scam and telemarketing calls increased from 2.14 billion a month in 2024 to 2.56 billion in 2025, so far. 

And automation is making them more effective. In 2025, a scammer used the AI tool Claude to automate attacks against at least 17 organizations. It was fully automated, and it could have been the most lucrative AI cybercriminal operation ever. 

Because AI has made hacking easier, so much more of our personal data is available to be used against us. It’s a perfect storm for AI to scam people. 

#2: Scam calls 

Generative AI can create deep fakes that realistically impersonate people’s voices, and then use them for scams.

Deep fakes are often eerily accurate and, therefore, unsettling. 

For example, in 2023, a mother received a phone call from what sounded like her daughter. Her daughter said she had been kidnapped and needed money wired to the caller in order to be freed. In a panic, the mother almost sent the money. But police officers suspected it was a scam. 

The family called her husband directly; their daughter was completely safe. The mother almost lost thousands of dollars. 

#3: Phishing text scams

While phishing text scams aren’t new, AI can make them more believable. 

AI databases collect large amounts of information about us and scammers can use this gathered data to create high-quality, personalized scams.

The content generated by AI can be so realistic that, when Harvard researchers created a phishing campaign using AI and tested it on Harvard students, about 60% of students in the study were fooled. 

#4: AI-generated identities

In “pig-butchering scams” (yes, the name is awful), scammers play the long game. Often, AI scammers or bots create fake identities online and interact with potential victims in hopes of gaining a victim’s trust.

Oftentimes, these scammers pretend to be romantically interested in their victims. By drawing on exposed personal data, they form their identities to appeal to the specific person. Eventually, they ask the victim to invest with them. 

In 2020, a California man received messages from a stranger on WhatsApp and struck up a conversation. Eventually, the stranger tricked him into scams that cost him over $1 million dollars. 

He’s not alone — thousands of other people have been victimized by pig-butchering scams. And, like scam calls and texts, bots can carry out more of these scams, and faster. 

How can we protect ourselves from AI scams?

First, be aware that scammers can use photos and videos of you that are already on the internet to create deep fakes of you. 

This doesn’t mean you should erase your online presence. But you should be careful of who you trust over the phone. Are they asking you for money? Does their voice sound different? Even if it doesn’t, does something seem off?

Consider deciding upon a safe word with loved ones to know if it’s really them talking to you.

And, be wary of calls or pop-ups and messages you see on your devices. A few red flags to look out for include anything that says you’ve made a large amount of money, interactions that ask you to invest or open accounts, or anything advertised as “the next big thing.” Overall, you shouldn’t comply with any unexpected request for information or money on the spot. If you think it’s legit, verify the request independently. 

Can we stop AI from making scams worse?

Despite the serious financial havoc that scammers can wreak, there hasn’t yet been any comprehensive attempt to regulate AI in the U.S.

Phone companies aren’t doing enough to protect us either. In our Who’s Calling? report, half of the largest phone companies received D or F grades on robocall protections. 

Phone companies should be doing more to protect us from scams. That’s why we’re urging them to adopt practices that prevent people from falling victim to scams. Many of these changes are free and feasible, like automatically providing on-screen scam warnings for incoming calls or allowing customers to block calls that don’t display Caller ID. 

And because AI scams are becoming more sophisticated, we’re also urging the Consumer Financial Protection Bureau to protect consumers’ data from scammers, hackers and other threats. Tell the CFPB to take action now.