The cybersecurity sector is experiencing a significant transformation, with funding for agentic AI startups expected to hit $3.8 billion in 2024, up from $1.3 billion the previous year. This increase signals a shift within the tech community, moving from generative AI to agentic AI. However, this evolution brings various security challenges that industry leaders are beginning to address.
The rapid development of AI agents, which can perform complex tasks independently, raises important questions about their function in cybersecurity. Savill, a notable figure in the field, emphasizes that traditional security measures are no longer adequate. “For years, security was intended to be the wall that kept the threats out, and resilience was what happened after a breach,” he states. “Today, true security means withstanding and recovering from attacks with minimal business impact.” This new approach requires not only enhanced defenses but also a solid governance framework for AI technologies.
The Rise of Agentic AI
Agentic AI sets itself apart from its generative predecessors by shifting from command-based operations to goal-oriented tasks. Rather than just following commands, these AI agents are assigned objectives, like infiltrating a company’s systems to extract sensitive financial data. Savill warns that this ability significantly increases the risks tied to AI deployment. “With agentic AI, you don’t give it a command — you give it a goal,” he explains. “The AI agent is going to perform its task, building plans and adapting as it goes.” This raises concerns about potential misuse, particularly in cyberattacks.
A recent Team8 CISO survey shows that over a third of cybersecurity leaders now consider securing AI agents their top challenge. As organizations integrate these autonomous agents into their operations, establishing stable governance mechanisms becomes essential. Savill advocates for “governed innovation,” stressing the need for regulatory frameworks that secure AI systems alongside traditional infrastructure.
The Imperative for Governance
Governance is crucial for the safe integration of AI technologies. Savill asserts, “Governance is the foundation for a safe scale.” He argues that implementing structured oversight early in the deployment process can help organizations leverage agentic AI's potential while effectively managing risks. This proactive strategy allows businesses to set the ‘rules of the game’ before AI agents become essential to their operations.
To enhance security, Savill recommends adopting a secure-by-design approach in the development of AI agents. Utilizing a policy engine can ensure that organizations are prepared to contain potential fallout from unforeseen circumstances. “A secure-by-design approach to creating AI agents can mitigate the risk of agents making up their own rules,” he notes, highlighting the importance of foresight in governance strategies.
As the agentic AI sector continues to grow, the need for effective governance and security measures will become even more pressing. Cybersecurity leaders must stay alert, adapting their strategies to incorporate these advanced technologies while protecting their organizations from emerging threats. The balance between innovation and security will shape the future of AI in cybersecurity, making it essential for enterprises to prioritize governance alongside technological progress.
The rapid expansion of agentic AI brings both opportunities and challenges. By adopting a governance-first mindset, organizations can navigate this changing landscape, ensuring they reap the benefits of AI while minimizing risks. Ongoing discussions about AI governance will be vital as the industry adjusts to a reality where autonomous agents play an increasingly significant role in cybersecurity.
The stories that move AI & crypto markets — before the market reacts.
Free. 7am ET. Five stories. 62,400 readers.
