Artificial Intelligence & Machine Learning
,
Events
,
Identity & Access Management
Barracuda’s Rohit Ghai on Agent Sprawl and the Need for Safer AI Security
The move from tool sprawl to agent sprawl is fundamentally changing the risk profile of cybersecurity. Fragmented tools once produced conflicting information, but fragmented agents now produce conflicting actions, with far greater consequences.
See Also: Taming the Rise of Shadow AI Agents
“We are unleashing largely autonomous capability in cyber without fully understanding where that might go,” said Rohit Ghai, CEO at Barracuda. “Before focusing on speed and efficacy, you have to focus on safety and ease of use,” especially for organizations that lack the human capacity to absorb added complexity.
Ghai said vendors introducing artificial intelligence-powered security tools must be challenged to demonstrate that their own AI has been hardened against prompt injection, data leakage and shadow AI risks before those tools are deployed.
In this video interview with Information Security Media Group at RSAC Conference 2026, Ghai also discussed:
- Why incident response playbooks built for pre-agentic threats are no longer adequate;
- How a risk-based approach for identifying crown jewels before applying protection underpins resilience;
- Why guardrails for agentic AI remain unproven and the boundaries of autonomous action are still undefined.
Ghai leads Barracuda’s strategy and growth, drawing on more than 20 years of leadership experience in cybersecurity and SaaS. He has expertise in digital identity, data privacy, AI security and risk management. Before joining Barracuda, he was CEO and board member of RSA Security.
Click Here For The Original Source.
