Updated on February 10, 2026 • 4 min read
Genesys Debuts an “Industry-First” Contact Center Virtual Agent, Leveraging Large Action Models
Director of Content & Market Research
February 10, 2026
Genesys has unveiled a new virtual agent solution, which it claims to be an “industry first.”
The all-new Genesys Cloud Agentic Virtual Agent harnesses large action models (LAMs), rather than large language models (LLMs) alone.
LAMs are purpose-built to predict next-best actions, not words. They are also smaller than most LLMs, designed to execute tasks with greater speed and accuracy.
Per 2025 Salesforce research, LAMs have matched or outperformed top LLMs - including DeepSeek, ChatGPT-3, and LLama - while being significantly smaller in size.
Consequently, the all-new Genesys Cloud Agentic Virtual Agent may be the beginning of a broader trend of customer-facing AI solutions utilizing this technology to improve performance and reduce consumption costs.
In terms of boosting performance, LAMs chiefly aim to improve the reasoning of virtual agents.
As a result, agents may better determine how to resolve a customer’s issues, leveraging company knowledge and customer data, alongside journey and process maps stored within Genesys’ orchestration software.
From there, the agent can adapt its approach and take action across the contact center environment to resolve more interactions autonomously.
In this sense, Genesys hopes to pivot from finding answers to completing tasks, which is crucial for boosting self-service success rates. Per 2025 Gartner research, these sit at just 22% across the customer support space.
Further in the future, Genesys’ agents may also collaborate with AI agents in enterprise systems beyond the reach of contact center representatives, automating longer-tail resolution flows.
While agent-to-agent communication, through Agent-to-Agent (A2A) and Model Context Protocol (MCP), has some way to go, it's an exciting prospect, which Genesys hopes to lead.
“With our LAM-powered Agentic Virtual Agent, we’re enabling AI to reason, plan and safely take action across systems. This gives organizations a responsible way to move beyond conversations and deliver consistent outcomes customers can rely on.”
Customers can access, configure, and design agentic virtual agents via the Genesys Cloud AI Studio, where they may also define permissions, guardrails, and behaviors aligned with company best practices and policies.
As they do so, Genesys suggests customers will likely experience fewer hallucinations and escalations to live reps.
Yet, for Hayley Sutherland, Research Manager for Conversational AI at IDC, the real breakthrough in this announcement is the solution’s potential to help legacy AI customers move beyond scripted, turn-based solutions that primarily exist to exchange information.
“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.”
Many of the 650 conversational AI vendors selling to customer service and sales teams (per recent estimates) may make similar promises about leveraging AI agents to complete multi-step resolution workflows.
However, the inclusion of LAM models, trained for customer service use cases, helps set Genesys apart.
Such thought leadership is crucial in a crowded space where differentiation is hard to deliver beyond marketplaces, ecosystems, and execution.
Moreover, it may give Genesys’ deep install base the confidence to work with their contact center provider on customer-facing agents, rather than a pure-play conversational AI player.
Notably, that install base is deep, with Genesys becoming the first tech provider to surpass $2 billion in annual recurring CCaaS revenue last year.
An Alternative Take: What's Most Interesting Is Genesys's Approach to Governance
While Genesys may be the first vendor to actively incorporate LAMs, many others are bridging the gap between conversation and action.
Additionally, they will likely soon release solutions that leverage LAMs, given the rapid pace of change in AI.
Ian Jacobs, VP and Lead Analyst at Opus Research, makes this point, yet spotlights something more differentiative in Genesys's approach.
"What’s most interesting here is that Genesys is intensely focused on governance, risk, and compliance from the start," he told CX Foundation.
"We have definitely seen a large gap in enterprise governance readiness for agentic AI. Many, probably most, organizations lack comprehensive policies about agentic AI or ones that lend themselves to govern the actions of agentic AI."
"Genesys explicitly calls out built-in action-level explainability, transparent decision paths, and audit trails as core parts of the LAM-driven virtual agent."
"Typical LLM governance discussions focus on text generation safety (things like bias control and content filters) rather than action traceability. If the agent can indeed take action, actually explaining why decisions and actions were taken and providing audit artifacts for compliance, that becomes the key to scaling success."
"For now, at least, the differentiator for vendors is less “better AI” and more “better accountability,” concluded Jacobs.