'AI is better for cyberattacks rather than cyberdefense'
By Wang Xueying
“A hack can happen to anyone…nobody will be safe from Internet attacks," even though people could develop artificial intelligence (AI)-powered software to help defend them, Heather Adkins said on Monday, director of information security and privacy, as well as a founding member of Google's security team.
As a cybersecurity specialist who has been responsible for protecting Google’s systems for over 15 years, Adkins pointed out that consumers should take care when protecting their own personal information during social communication on the Internet, especially via emails.
According to Google, more than one billion people in the world use Gmail to communicate online. While more AI-powered security programs have been developed to protect the digital privacy of netizens, AI-powered software has failed to be particularly effective at preventing “even 1970s-era attack methods” from happening, let alone recent hacks using advancing new technology, said Adkins.
In his opinion, AI technology is better for cyberattacks than cyberdefense because “it produces too many false positives”.
In general, an AI machine needs to “learn what is good and bad” by analyzing various feedback. However, “we’re not sure which one is good and which one is bad,” especially when hackers are able to "mask" the true nature of malware, she said.
What could we do to keep our networks safe? “More talent…less technology,” said Adkins.