Agentic AI Arms Race Heats Up: Enterprises Scramble for Defense

The agentic AI arms race is heating up in the cybersecurity world, with enterprises facing new challenges in defending against autonomous attacks. According to a recent poll by Dark Reading, 48% of security professionals identify agentic AI and autonomous systems as the most dangerous attack vector.

Salesforce's security leaders agree that velocity is the key factor in this threat landscape. 'Finding vulnerabilities has never been the hardest part of the job,' said Iain Mulholland, Chief Information Security Officer at Salesforce. 'The difficult task is fixing them rapidly, reliably, and at scale.'

To combat this, Salesforce has developed a proactive approach to AI defense, with human oversight playing a crucial role in preventing overcorrections. 'We have to be right every time,' said Kelly McCracken, Deputy Chief Information Security Officer and SVP of Security Operations at Salesforce.

The company's Customer Zero posture involves treating itself as its first and toughest customer, using its own technology to identify gaps early on. This approach has allowed Salesforce to develop a vulnerability agent that can close vulnerabilities in under an hour, compared to the seven-day runway previously available.