AI agents are supposed to make work easier. But\u00a0they\u2019re\u00a0also creating a whole new category of security nightmares.\u00a0<\/p>\n
As companies deploy AI-powered chatbots, agents, and copilots across their operations, they\u2019re facing a new risk: How do you let employees and AI agents use powerful AI tools without accidentally leaking sensitive data, violating compliance rules, or opening the door to prompt-based injections? Witness AI just raised $58 million to find a solution, building what they call \u201cthe confidence layer for enterprise AI.\u201d\u00a0<\/p>\n
Today on TechCrunch\u2019s Equity<\/a> podcast, Rebecca Bellan was joined by Barmak Meftah<\/a>, co-founder and partner at Ballistic Ventures, and Rick Caccia<\/a>, CEO of WitnessAI, to discuss what enterprises are actually worried about, why AI security will become an $800 billion to $1.2 trillion market by 2031, and what happens when AI agents start talking to other AI agents without human oversight.\u00a0<\/p>\n Listen to the full episode to hear:\u00a0\u00a0<\/p>\n How enterprises accidentally leak sensitive data through \u201cshadow AI\u201d usage.\u00a0<\/p>\n What CISOs are worried\u00a0about right now, how the problem has evolved rapidly over\u00a018 months, and what it will look like over the\u00a0next\u00a0year.\u00a0<\/p>\n Why\u00a0they think traditional cybersecurity approaches\u00a0won\u2019t\u00a0work for AI\u00a0agents.\u00a0<\/p>\n Real examples of AI agents going rogue, including one that threatened to blackmail an employee.\u00a0<\/p>\n Subscribe to Equity on YouTube<\/a>, Apple Podcasts<\/a>, Overcast<\/a>, Spotify<\/a>, and all the casts. You also can follow Equity on X<\/a> and Threads<\/a>, at @EquityPod.<\/p>\n