Infosys: Embed Responsible AI & Security in Agent Lifecycle

Infosys executive Balakrishna DR, or Bali, says organizations need dynamic governance for AI as technology and regulations rapidly evolve. He emphasizes embedding responsible AI, risk management, and security throughout the agent lifecycle. What's interesting is that Infosys has established a dedicated inter-disciplinary body, the Infosys Responsible AI Office, to address these challenges. Bali notes that a major barrier to scaling AI is context and data readiness, as most enterprise data isn't structured or trusted for autonomous agents. He also highlights "governance debt," where pilot projects often skip necessary controls, making retrofitting difficult. This makes the role of System Integrators even more crucial. Bali observes that human-technology interaction is shifting from clicks to intent, with people describing outcomes and multi-agent systems planning and executing. The bottom line is that enterprises are redefining work, moving from tasks to outcomes and from static software to adaptive systems, with humans focusing on judgment and oversight while agents handle repeatable execution. This matters because it outlines the critical steps for safely and effectively integrating AI into businesses.