Meta on Wednesday warned that hackers are using the promise of artificial intelligence like ChatGPT to trick people into installing malicious code on devices.
Over the past month or so, the social media giant’s security analysts have found malware masquerading as ChatGPT or similar AI tools, chief information security officer Guy Rosen said.
“The latest wave of malware campaigns that have noticed generalist AI technology are capturing people’s imaginations and excitement,” said Rosen.
Meta, the parent company of Facebook, Instagram and WhatsApp, often sharing what they learn with industry colleagues and others in the cyber defense community.
According to Rosen, Meta has seen “threat actors” favor internet browser extensions that promise broad AI capabilities but contain malware designed to infect devices. .
In general, hackers often set their traps with attention-grabbing developments, tricking people into clicking web links with traps or downloading data-stealing programs.
“We’ve seen this on other popular topics, such as crypto scams driven by tremendous interest in digital currencies,” Rosen said.
“From a bad actor’s point of view, ChatGPT is the new cryptocurrency.”
According to the tech company’s security team, Meta found and blocked more than a thousand web addresses that were advertised as promising tools like ChatGPT but were in fact traps set by hackers.
Rosen said the meta has yet to see generalized AI being used more as a decoy, but is bracing for the inevitable possibility that it will be used as a weapon.
“Creative AI holds a lot of promise and the bad guys know it, so we should all be extra vigilant to stay safe,” Rosen said.
At the same time, Meta teams are working on general AI uses to combat hackers and deceptive online influence campaigns.
“We have teams thinking about how (next generation AI) can be abused and the defenses we need to put in place to combat that,” said Meta’s head of privacy policy , Mr. Nathaniel Gleicher said at the briefing.
“We are preparing for that.”
