ChatGPT is getting pretty good at writing phishing emails, according to new IBM research.
Why it matters: Cybersecurity experts have long warned that hackers could weaponise ChatGPT and similar AI tools to scam unsuspecting targets.
Driving the news: IBM researchers recently ran an experiment among an unspecified global health care company’s roughly 1,600 employees.
Half of the employees got a phishing email written by IBM cybersecurity experts. The other half got one made using ChatGPT.
By the numbers: 14% of employees who received the human-written phishing email fell for it and clicked on a malicious link.
But the ChatGPT-written email was pretty close, with an 11% click rate.
It only took five minutes for Stephanie Carruthers — IBM’s chief people and experiment leader — and her team to create the ChatGPT email.
Her team usually needs about 16 hours of research and prep time to write a believable phishing email.
Be careful what you respond to people 👀
