How AI is Changing the Cyber Threat Landscape

Key Takeaways:  

• Advanced AI has enabled new scams such as deepfakes and voice cloning calls, while also making traditional scams like phishing emails far more convincing.
• Most scams still share common red flags, including false urgency, unusual payment requests and inconsistencies in details.
• You can protect yourself by slowing down, verifying through trusted sources and reporting suspicious messages or calls to your bank.

Artificial intelligence (AI) is reshaping daily life. It powers the voice assistants in our homes and the cars that can drive themselves, but this technology is also being exploited by cybercriminals. Scammers are now leveraging advanced AI tools to supercharge their schemes, producing personalized emails, videos and even voice messages that look and sound remarkably real. 

We are in a cyber threat landscape where the old warning signs aren’t always enough. To stay safe, consumers need to recognize how AI-driven scams work and the steps they can take to protect themselves. Let's dive in.

Deepfakes

Deepfakes are videos or images created with AI that make it appear as though someone said or did something they never actually did. Cybercriminals can use AI to impersonate public figures, trick people into financial scams or spread false information.

Because the visuals and audio can seem realistic, deepfakes can easily dupe viewers who don’t know what to look for.

How to spot a deepfake

Even high-quality deepfakes often contain subtle flaws. Watch closely for:

• Unnatural facial movements such as mismatched lip syncing, stiff expressions or odd blinking patterns.
• Lighting or shadows that don’t line up, especially if the subject looks edited into the scene.
• Audio that feels slightly off, with a flat tone or mismatched timing between voice and mouth movements.
• Strange background details, like warping or blurring around the face or edges of the video.

What to do if you suspect a deepfake

If you come across a video that seems suspicious, the safest approach is to slow down and verify before reacting. Start by checking the source: If the clip only appears on a random account and not on a trusted news outlet or the official page of the person involved, that’s a major red flag.

Consider cross-checking the story with reliable coverage as well, and avoid sharing the video until you’re certain it’s authentic. If you still have doubts, report the content through the platform’s tools or contact the organization directly to confirm whether it’s real.

Phishing emails

Phishing has been around for years, but AI has made it far more convincing. Instead of clunky messages full of spelling mistakes, today’s scammers use AI to write emails and texts that are polished, professional and even personalized for your inbox.

These messages often appear to come from a bank, delivery company or government agency, urging you to click a link or provide sensitive information. 

How to spot a phishing email (or text)

While AI-crafted phishing emails can look real, there are a few signs that give them away:

• Scammers can’t send messages from any verified accounts, so check if the sender’s address is slightly misspelled or from an unfamiliar domain. For example, "bank" might be misspelled as "bamk."
• Many phishing attempts use generic greetings like “Dear customer” instead of addressing you by name, even when legitimate companies normally personalize their messages.
• Unexpected attachments or links are another warning sign, and hovering over the link often shows a web address that doesn’t match the company it claims to be from.
• These emails often create a false sense of urgency, warning that your account will be locked or suspended unless you act immediately.

What to do if you suspect phishing

If an email or text feels phishy, the best step is to avoid clicking any links or downloading attachments, since these could expose you to malware or fake websites. Instead, type the company’s web address directly into your browser or call the organization using a phone number from its official website, not the one provided in the message.

Most email providers offer a “report phishing” option, and many banks encourage customers to forward suspicious messages for review. Once you’ve reported it, delete the email so you don’t accidentally interact with it later.

Voice cloning calls

With only a few seconds of recorded speech (which can be taken from a social media video, voicemail or online clip), scammers can use AI to create a near-perfect replica of someone’s voice. They typically use these cloned voices to impersonate a loved one in distress, a coworker making a request or even a bank representative asking for account information.

These systems use advanced AI, similar to the chatbots many people already interact with, to recognize speech and generate realistic responses in real time. That means a scammer’s AI can answer simple questions, carry on basic conversation and even adapt if you interrupt or ask for clarification.

How to spot a voice cloning call

Recognizing these subtle signs can help you stay calm and avoid falling for a convincing imitation:

• Cloned voices often have unnatural pacing, with odd pauses or speech that doesn’t flow like a normal conversation.
• The tone of voice can sound flat or inconsistent, lacking the natural emotion and variation you’d expect from the real person.
• Some calls may sound suspiciously “clean,” with no background noise or ambient sounds you would normally hear.
• The caller might avoid answering specific questions, repeat vague phrases or steer the conversation back toward urgency or payment.

What to do if you suspect a voice cloning call

If you ever feel uneasy during a call, trust your instincts and pause before responding. Hang up and call the person back on the number saved in your contacts. You can also test the caller by asking a personal question that only the real person would know.

If the caller is pushing for secrecy, urgency or unusual payment methods, treat it as a red flag and verify the request with another trusted family member, friend or colleague before taking any action.

Scam prevention checklist

AI may be making scams more convincing, but the fundamentals of protecting yourself haven’t changed. Most scams, AI-generated or not, have similar warning signs. Before you click, reply or send money, it helps to pause and ask a few simple questions:

• Am I being pressured to act right now? Scammers often create urgency so you don’t have time to think.
• Is the payment method unusual? Requests for gift cards, cryptocurrency or wire transfers are a common sign of fraud.
• Do the details check out? Look carefully at email addresses, phone numbers and links. Do they match what you know to be legitimate?
• Why does this message feel emotional? Fear, guilt or excitement can be signs you’re being manipulated.
• Does this sound too good to be true? Offers with guaranteed returns or surprise rewards almost always are.

Pausing to ask these simple questions can help you cut through even the most realistic AI-generated scams. Remember, your bank will never pressure you to act immediately, demand secrecy or ask for unusual forms of payment like gift cards or cryptocurrency.

If you receive an unexpected message from someone claiming to be your bank, do not click on the link or call the number. Instead, call your bank to report. 

To learn more about how to protect yourself online and spot potential cyber threats, visit our Security Center.