Google Labs Upgrades Opal With AI Agent Step Feature, Enabling Dynamic Workflows Without Pre-Defined Paths

Google Labs upgrades Opal with a new 'agent step' feature powered by Gemini models, enabling dynamic, goal-driven AI workflows with persistent memory, natural language routing, and human-in-the-loop orchestration — signaling a major shift in how enterprise teams can build and deploy intelligent agents without rigid, pre-defined paths.

• Google Labs has updated Opal, its no-code visual agent builder, introducing an 'agent step' feature that enables dynamic, goal-directed workflows powered by Gemini 3 models, allowing agents to select tools, route decisions, and interact with users without pre-defined paths.
• The update establishes three core enterprise agent capabilities: persistent memory across sessions, dynamic natural language-driven routing, and human-in-the-loop orchestration, where agents autonomously determine when to pause and seek human input rather than relying on hard-coded checkpoints.
• Enterprise teams are being signaled that foundational AI agent patterns, including adaptive planning, memory management, and dynamic routing, are now productized and accessible, urging IT leaders to move away from over-constrained architectures and embrace goal-driven, model-managed agent design.