Artificial intelligence continues to reshape the enterprise landscape, but many organizations are still determining how to implement it responsibly, securely, and at scale. At this year’s Enterprise Connect 2025, Zoom’s Leo Bolton, Head of Product, Solutions and Industry Marketing and Patrick Kelley, Chief Technology Evangelist shared how the company is advancing its AI capabilities and assisting enterprise customers in adopting the technology.
The Evolution of Zoom AI Companion
Zoom’s AI journey began two years ago with the launch of AI Companion, initially offering meeting summaries and productivity enhancements.
Since then, the platform has evolved dramatically. In the most recent announcement, Zoom unveiled agentic capabilities—a new class of intelligent, task-performing agents integrated throughout its ecosystem.
“We started with meeting summaries, and then we expanded to Zoom AI Companion 2.0,” said Leo Boulton. “Now, we’re talking about agentic skills deployed across the entire platform—not just for meetings, but across contact centers, workflows, and more.”
‘Agentic‘ is the latest marketing phrase being thrown around by an industry that is still new to the game. Some vendors are quite ‘loose‘ using the agentic moniker, so users must be careful in understanding what they are buying.
So, what is agentic AI? Agentic AI is artificial intelligence systems that can autonomously perform tasks, make decisions, and interact with users or systems in a goal-oriented way—essentially acting as intelligent agents. Unlike traditional AI tools that assist with single tasks, agentic AI can initiate actions, adapt to changing inputs, and carry out multi-step processes with minimal human intervention. In the enterprise context, this means AI supports workflows and actively drives them—handling customer queries, summarizing meetings, or resolving IT issues.
Zoom introduced 45 specific agentic skills at launch and hinted at many more in development. These agents don’t just assist—they act. For instance, a virtual agent in the contact center can autonomously handle routine inquiries, freeing human agents for more complex work. The key word is ‘autonomous‘ here; Zoom brings true agentic solutions.
Tailoring AI for Enterprise Needs
The industry has thus far coalesced around some fairly horizontal mass use cases for AI. For UC Today, this has led us to a deep dive into how it drives the meeting room and how contact center agents have had their roles augmented by the technology.
However, the next phase of AI, or agentic AI, will revolve around customization and building models that perform unique operations or deliver unique value to verticals.
Zoom is one of a handful of vendors that has identified this need and the key advantages of customization. While out-of-the-box functionalities like meeting summaries are driving value around meetings, Zoom’s agentic capabilities will not only allow for custom agent development but also integrate with other notable vendors.
“Zoom is an open platform,“ Leo emphasized. “It doesn’t live in isolation—it integrates deeply with Google, Microsoft, and other enterprise ecosystems.”
This openness means businesses don’t need to overhaul their entire infrastructure to benefit from AI. Instead, they can layer Zoom’s capabilities into their existing workflows. For customers, this makes the benefits of Zoom’s AI infinitely easier to adopt. For partners, this gives them new capabilities to master and can bring the benefits of AI quickly.
Patrick Kelley added, “The fact that we can stretch Zoom AI Companion into a Microsoft or Google world is what drives real enterprise value.”
The importance of these integrations cannot be underestimated. They will deliver key benefits for users who have adopted Zoom for single-use cases like meetings.
Alleviating Adoption Fears
Data privacy, control, and compliance are major hurdles for enterprises adopting AI. I spoke with end users at Enterprise Connect, and the common theme was that there are a lot of POC and experimentation going on, but nothing is making it into Zoom to address these fears directly.
“Our promise has always been responsible AI,“ said Leo. “That means human-in-the-loop oversight, full customer control, and no use of customer data to train models.”
Zoom has adopted a federated AI approach and implemented administrative controls, ensuring compliance and data protection across the board. In addition, third-party integrations like Theta Lake provide additional compliance and security features tailored to regulated industries.
Patrick highlighted, “We’ve built compliance tools like our integration with Theta Lake that give IT managers the ability to put real guardrails around AI data use.”
Getting Started
For organizations just beginning their AI journey, both Leo and Patrick gave clear advice: start small and think practical.
“You don’t have to boil the ocean,“ Patrick said. “Start with simple workflows. Meeting summaries, for example, are a great entry point. Without user intervention, you get immediate value—automatic recaps, action items, follow-ups.”
This approach allows teams to experience AI benefits firsthand without large-scale disruption while building momentum for more complex implementations later.
Zoom is onboarding new technology for a new generation of workers entering the workplace. Understanding how Gen Z will use technology and where they will be able to add value will be crucial for businesses to understand moving forward.
Aside from including tools to help IT teams adopt the technology, Zoom is also leveraging the power of its partners or trusted advisors to drive understanding and adoption across its customer base. For IT teams, AI is just too compelling to ignore. The potential upside of getting AI right is justification enough to experiment, test, and reap the rewards.
For a full rundown of the Zoom Workplace platform, see our Ultimate Guide.