LAS VEGAS — The cybersecurity industry’s obsession with generative AI continued for a second year in a row at Black Hat 2024, where security experts are trying to figure out if the revolutionary technology is friend or foe.
“The opinion is split down the middle,” said Audra Streetman, a security strategist at Cisco Systems Inc.’s Splunk division, citing a recent survey of 1,600 security executives who believe AI is the solution, as well as probable cause, for potential breaches.
“It is a double-edged sword,” added Brett Stone-Gross, senior director of threat intelligence at Zscaler Inc. “There is a lot of hype, and we need to separate the hype from reality. Some look to LLM as a solution, which is good at some things but can also have bad answers.”
What makes the schizophrenic attitude toward genAI among executives and security experts is the speed in which it is being applied while security moves at a more leisurely pace. Dave DeWalt, founder and chief executive of cyber-focused VC firm NightDragon, believes genAI has a two-year head start on security defenses, leading to a broadening attack surface.
“Security is friction to genAI [adoption] speed,” Leo Scott, chief innovation officer at DataTribe, said in an interview. “But security is the tail of the dog.”
And as the pace of deployment quickens, the attack surface has broadened, adds Operant AI CEO Vrajesh Bhavsar.
One intriguing angle to the raging love-hate debate centers on whether AI, which has revolutionized computing, is leaving the more incremental pace of security innovation in the dust. The technological contradiction has forced many tech chief executives, in their zeal to quickly pivot operations to AI, to ponder how to avoid strangling their security teams?
At Qualys, the ticklish balancing act fell to CEO Sumedh Thakar, who saw the opportunity to leverage AI to super-charge operations as well as bolster cybersecurity and spent accordingly. “What we experienced was not very different from adopting the cloud, and making it scalable without sacrificing security,” he said in an interview. “There is risk involved, and you need to mitigate it.”
On Monday, Qualys expanded its portfolio with Qualys TotalAI to address the risks associated with securing genAI and large language model applications.
Indeed, as executives warm to safe AI use cases, they are increasingly likely to accelerate its implementation — especially as regulation of AI gains steam through agencies like the Securities and Exchange Commission, according to Stone-Gross.
“A year ago, AI was the shiny object, but now people are warming to mission-critical applications, just as they did with the cloud and online banking before it,” Coalfire Systems Inc. CEO Tom McAndrew said in an interview. “Like any new technology, it is good and it is bad. But without AI, you’re considered a dinosaur.”
