The man who uncovered a major celebrity data leak has a bleak warning about the ‘super scary’ capabilities of AI, and it has nothing to do with losing your job.
An ethical researcher with 15 years experience, Jeremiah Fowler, has found significant data breaches at companies and governments and seen the damage cybercriminals can do if they get their hands on sensitive and personal information.
The rise of artificial intelligence and the computing power it has given to those cybercriminals is now really worrying him.
It’s not just institutions that are at risk, but all of us, as anyone could fall for a scam at the best of times, let alone now that their creativity and sophistication has been increased ‘100-fold’ with AI.
That’s making scams much tougher to spot, with Fowler warning an ‘AI turd tornado’ is coming.
“AI has given cyber criminals steroids,” Fowler tells LADbible.
“How many emails did you used to get about winning the Nigerian lottery? You only need to send $500 in gift cards and all the money is coming to you?
“You can read that, the broken English, the grammatical mistakes. You knew in 30 seconds it was a scam. Now, scammers can use AI that’s incredibly targeted.”
Cybercriminals have a big advantage with AI (Getty Stock)
How AI has powered criminals
Scams have always existed. From confidence tricks played by a mystery rogue who rides into town to the hoards of fake ticket sellers, phishing expeditions and catfishers that plague the internet today.
However, AI has given very powerful tools to even non-technical people that wouldn’t have had the ability to pull off scams just a few years ago, Fowler warns.
It’s especially concerning as it allows scammers to process remarkable amounts of information in the blink of an eye, something that would have taken weeks or months without the technology.
As a result, Fowler says criminals can use information already out there on the internet and social media – just search your own name and see what the results are and try not to panic – coupled with any historically breached data that AI can find to ‘build a shockingly accurate target profile.’
By using AI to scrape information from previous data breaches, criminals could find if yours or anyone you know’s passwords have already been leaked, access accounts and then use personal information to create bespoke social engineering scams on you or your loved ones.
“As the creativity of criminals is combined with AI, we will see more risks in terms of phishing and social engineering being more believable and attacks more successful,” Fowler said.
Scams could become harder to spot (Getty Stock)
Phishing exams could become much harder to spot
A phishing scam is designed to trick a person into clicking a link which contains malware and can be used to steal information or cripple systems.
They can already be sophisticated, looking like genuine emails from your boss or line manager to dupe you into clicking a nefarious link without even realising.
They’re already a massive problem companies have to deal with regularly, and you have probably been given training in spotting them if you work on computers and handle sensitive data.
With AI, they’re only going to get better.
Fowler said: “Right now, AI-powered phishing is the biggest risk that I see because of how good it can compile personalised data that is specific to that individual from hundreds or thousands of data points around the web and beyond.
“Soon AI will know more about you than you know about yourself.”
Another technique that’s deployed is slightly changing email addresses so a person doesn’t realise they’re communicating with an impostor.
Fowler uncovered a major data leak which included contact information from a host of Hollywood superstars, and believes this technique could have been used if any ‘bad guys’ for their hands on an email address.
He explained: “I see a famous actor, I see the email address, I just change it by one letter and then I email all the people they may have contact with.
“I’m like ‘hey, can you give me an update on this? Oh, check out this PDF file,’ that just happens to have malware.”
Anyone could fall for a scam (Getty Stock)
Deep fake scams could be a huge problem
AI has led to the advent of deep fakes, which are realistic looking and sounding replications of a person’s appearance that can be deployed in real time, a digital mask over a scammer’s face, not just in recorded videos.
There have already been major deep fake scams, with finance workers being fooled into handing over millions of dollars, and the technology is getting better.
Though they are in the early stages of mass adoption by criminals, Fowler believes the use of deepfakes, voice cloning and synthetic identity fraud are going to be a big danger.
He said: “We are all at risk when these AI scams become the standard and grandparents and parents are receiving regular calls and videos from criminals who have cloned their loved ones, friends, family, or business contacts demanding money.”
Thankfully, Fowler does have a handy trick he already deploys to try and protect himself from any would-be imposers.
” We are in the matrix where we can no longer believe what we hear or what we see and a healthy amount of paranoia will be needed to protect ourselves,” he said.
“I actually have a password with my inner circle. If someone doesn’t know the code word, I will end the communication.”
