Savinay Berry, CTO of OpenText, detailed his return to the company to lead AI integration, emphasizing OpenText’s extensive but often unseen role supporting essential systems for Fortune 5000 companies. He stressed the unique challenges of implementing AI in the enterprise, particularly concerning security and compliance.
The discussion centered on the promise of AI agents to proactively enhance business workflows. Savinay introduced the major risk of “agent sprawl”—the operational hazard posed by unregulated AI agents—and stressed the necessity of a secure platform to govern agent behavior and data access.
He also addressed AI security, comparing threats like prompt injection to traditional database vulnerabilities and acknowledging that AI advancements are rapidly increasing cybersecurity complexity. He cautioned against overestimating AI’s current impact on coding quality, advocating for maintaining strong engineering fundamentals and promoting open communication between vendors and partners to ensure responsible and collaborative AI development.
