Genesys®, a global cloud leader in AI-Powered Experience Orchestration, announced the industry’s first agentic virtual agent built with large action models (LAMs) for enterprise CX, enabling autonomous, end-to-end resolution of customer requests. The new Genesys Cloud™ Agentic Virtual Agent understands customer goals, determines the next steps and executes complex actions across front and back-office systems and teams. The launch marks a shift from reactive self-service to autonomous, outcome-driven customer engagement that reduces effort, improves operational efficiency and scales reliably across the enterprise.
Today, ambitions for bot-driven self-service are high, but confidence is still low. According to a Gartner® report, “The average self-service success rate is just 22%. Forty-six per cent of leaders surveyed indicated that increasing self-service success is a Top 3 priority for them in 2026.”
The challenge is that most automation isn’t built for the way people want to interact or how companies manage work. Many organisations have adopted virtual agents built on large language models (LLMs) because they significantly improve conversational quality. However, LLM-based virtual agents weren’t designed to execute multistep workflows that span systems or handle changing customer needs, leading to inconsistent outputs or breakdowns as complexity increases.
Genesys Cloud Agentic Virtual Agent closes this gap by pairing LAMs with enterprise-grade orchestration and governance to enable safe, autonomous resolution at scale. Several leading organisations, including M&T Bank, Banco Pichincha, a global Fortune 500 healthcare company and a Fortune 50 North American retailer, are already exploring the new capability to advance automation and improve how customer requests are resolved across their operations.
Serving as the central orchestration layer for autonomous customer work, these smarter virtual agents can progress workflows across CRM, billing, service operations and other enterprise systems while adapting as conditions change. Embedded guardrails, unified data and transparent decision paths help ensure every action remains explainable, policy-aligned and predictable.
These capabilities are strengthened through the recent partnership between Genesys and Scaled Cognition. The Genesys Cloud platform now uses the Scaled Cognition APT-1 LAM, purpose-built for deterministic, action-grounded execution. This enables autonomous AI that can responsibly act across enterprise operations, allowing businesses to carry requests through to resolution, reduce hallucinations and restore confidence in self-service.
Customers will no longer have to navigate rigid menus or restart their requests as issues grow more complex. The interaction will shift from “finding the right answer” to “getting the job done.” The Genesys Cloud LAM-powered virtual agent understands customer context and intent to safely drive real outcomes in a single, seamless interaction to reduce dead ends, repeated explanations and unnecessary handoffs.
“Autonomy in customer experience only works when it’s built on trust, transparency and control,” said Mr. Olivier Jouve, chief product officer at Genesys. “With our LAM-powered Agentic Virtual Agent, we’re enabling AI to reason, plan and safely take action across systems. This gives organisations a responsible way to move beyond conversations and deliver consistent outcomes customers can rely on.”
Genesys Cloud Agentic Virtual Agent extends the platform’s governance-first approach with action-level explainability, auditability and continuous learning. Through Genesys Cloud™ AI Studio, organisations can design, configure and govern Agentic Virtual Agents by defining guardrails, permissions and behaviours aligned to their policies and requirements. This provides explainability into how decisions are made and executed, while the system improves accuracy, containment and orchestration, enabling enterprises to scale autonomous engagement without sacrificing oversight or compliance.
Genesys plans to expand Genesys Cloud Agentic Virtual Agent through native support for open standards such as Agent-to-Agent (A2A) and Model Context Protocol (MCP), enabling secure collaboration with other AI agents and enterprise systems. This will allow organisations to orchestrate multistep, cross-system customer outcomes, maintaining shared context across interactions while applying centralised governance, policy enforcement and control.
Commenting on the news:
- “As enterprises push self-service further, the limits of scripted, turn-based interactions that can only support information exchange are becoming more apparent,” said Ms. Hayley Sutherland, research manager, conversational AI at IDC. “Resolving complex customer requests requires AI that can plan and execute multistep actions across systems, while remaining predictable and auditable. The focus of Genesys on combining autonomous execution with experience orchestration and governance reflects the direction the market needs to move to make self-service both more effective and trustworthy.”
- “At DNB, delivering seamless, trustworthy customer experiences at scale is a strategic priority,” said Mr. Trond Prestø, head of customer care at DNB. “We see agentic AI as an important next step in self-service by allowing us to resolve customer needs at scale. The way Genesys Cloud approaches autonomous virtual agents, with built-in controls and governance, aligns with how we think about responsibly modernising services while improving efficiency and consistency.”
- “In the enterprise, 80% accurate is 100% useless for automation,” said Mr. Dan Roth, co-founder and CEO of Scaled Cognition. “LLMs are mainly designed to generate text, not execute tasks—and in the real world, that gap leads to hallucinations and policy drift. The foundation of trustworthy automation is super-reliability, not super-intelligence. By combining our Large Action Models with the governance of Genesys Cloud, we will deliver a more reliable AI stack for enterprises around the world.”
Genesys Cloud Agentic Virtual Agent, powered by large action models, is expected to be generally available globally in the first quarter of the company’s fiscal year 2027 (1 February, 2026–30 April, 2026).
If you have an interesting Article / Report/case study to share, please get in touch with us at editors@roymediative.com roy@roymediative.com, 9811346846/9625243429.
