How AI is changing the face of cyber attacks

Cyber attacks are nothing new – for years businesses and customers have faced phishing attempts, spam emails and even malware and ransomware that try to pilfer their data and finances. But with the rise of AI, cyber attacks are changing.

At Everywhen, we want to help you stay one step ahead of cyber criminals. Read on to find out more about how they use AI in their tactics and what you can do to keep yourself and your business safe online.

It used to be relatively easy to spot most phishing emails – how often have you received spam with brand or website names spelled incorrectly? Misspellings and clunky wording are usually an instant giveaway. But unfortunately, AI can now be used to neaten all of this up, polishing grammar and spelling until there are no mistakes left1.

The AI can also generate entire emails for the scammer of varied length and structure, which  makes it even harder for email filters to see them as spam, as they won’t share the same template. Using content scraped from web pages means that these emails can be very convincing. That’s why it’s more important than ever to verify the sender before clicking on any links in any email or text you receive.

While it sounds like something straight out of The Little Mermaid, voice stealing is actually a real and documented phenomenon that can be used to extract money or data from customers or businesses. There are different versions of the scam:

Cyber criminals record audio of someone talking, either from an online source or over the phone, then use AI to create an audio “deepfake” – this deepfake is then used to manipulate people via phone call into giving up money or data.

Scammers use AI to clone customers’ voices and deceive banks and businesses into setting up unauthorised direct debit payments2

There are a few steps you can take to minimise your chances of falling prey to a voice stealing scam:

Install a spam filter on your phone so you can tell if someone’s calling you from an unknown or suspicious number.

If you’re not sure about what the person on the other end of the line is saying (and especially if they’re asking for money), hang up and call them back on a trusted number.

Sometimes these scams go even further and employ video deepfaking too, so it’s crucial to make sure you verify the caller via another channel before you do anything like transferring money or sensitive data.

Malware is any kind of program designed to damage or gain unauthorised access to computer systems. With the assistance of AI, these programs can become even more dangerous and hard to remove from computer systems – the AI continuously rewrites the malware’s code, which means that if it comes up against detection software, it can effectively disguise itself3.

Make sure your anti-virus and detection software are up-to-date and avoid clicking on any suspicious links, whether in emails or on the web.

This is when hackers hide AI prompts in otherwise innocent-looking websites or emails. The prompts are usually hidden by changing the font colour or size, and when an AI chatbot reads the prompt, it’s tricked into executing the command – which could lead to sensitive data being lost4.

Be careful about what websites you use and what emails you open – verify the addresses or sender before you click on anything or feed it into an AI chatbot.

Interested in cyber insurance for your business? Get in touch with Everywhen today on 0330 808 1880, or visit our dedicated cyber insurance page here.