AI-focused cloud service provider Akamai on Tuesday launched Agent Sandbox, a sealed environment for AI agents to run code without affecting a company's other systems. This move highlights a growing trend where AI agents are driving new cybersecurity spending, as software increasingly acts autonomously, posing risks of abuse by hackers or the agents themselves. Other major cybersecurity firms like Palo Alto Networks, Cisco, CrowdStrike, Okta and Zscaler are also offering protection for AI agents. A 2026 Palo Alto Networks survey found AI agents are expected to grow 85% over the next 12 months.