Editor’s Note:

This article is based on AiNews.com’s on-site coverage of Cisco’s WebexOne 2025 in San Diego, California. During the event, Cisco introduced major AI advancements that redefine collaboration, customer experience, and trust in enterprise AI. The insights below are drawn from keynote sessions, technical presentations, and exclusive interviews with Aruna Ravichandran, Cisco’s SVP and CMO of Collaboration, and Cullen Jennings, Cisco’s CTO for Audio, Video, and AI Collaboration.

Last week at WebexOne 2025, Cisco took its long-standing “Experience Matters” theme to a new level with the launch of Connected Intelligence, designed to facilitate collaboration not only between people but between humans and AI agents.

“Very soon the world will look like you have 80 billion digital workers,” said Aruna Ravichandran, SVP & CMO at Cisco. “So it’s no longer good enough to just focus on human-to-human collaboration. We’re continuing to go down the path of making experience really matter, but we’re taking it to the next step by facilitating human-to-AI and AI-to-AI collaboration.”

That shift signals a deeper transformation at Cisco — from enhancing meetings to redefining how humans and digital coworkers interact. Rather than building one all-powerful assistant, Cisco is focusing on a network of specialized AI agents that can coordinate across workflows, tools, and devices.

The company introduced a family of AI agents aimed at transforming meetings and daily operations:

“It gives you a cinematic moving experience,” Aruna explained. “You show up with your best foot forward — and that’s what experience is all about.”

Beyond the user-facing tools, Cisco also previewed AI Canvas for Control Hub, a visual dashboard that helps IT teams see how AI affects performance and reliability in real time. It’s part of a broader trend inside Cisco: blending creative AI experiences with the operational intelligence that keeps collaboration running smoothly.

While Connected Intelligence reimagines collaboration, Cisco’s Agentic AI for Webex Contact Center brings that same thinking to customer experience — where speed, empathy, and personalization often define success.

“We’re also launching AI Quality Management,” Aruna said. “This gives supervisors the ability to monitor the quality of conversations happening with virtual agents as well as live ones.”

By combining AI Quality Management (AI QM) with conversational analytics, Cisco is giving companies a clearer picture of sentiment, loyalty, and churn — in real time. The result is a contact center that doesn’t just respond faster but learns how to respond better.

AI QM is part of Cisco’s broader Contact Center AI Suite, which will be available to enterprise customers in early 2026. The company also announced expanded integrations with Salesforce, Epic, Amazon Lex, and Splunk, extending its reach across industries like healthcare, retail, and finance.

“We are announcing market-leading partnerships with Salesforce,” Aruna added. “And the entire focus is to help communities collaborate while keeping experience at the center — but taking it to the next level with our AI agents.”

Behind the scenes, these partnerships serve a larger goal: to make AI less of a standalone tool and more of an embedded layer of intelligence that supports every customer interaction.

If Connected Intelligence connects people and data, and Agentic AI connects businesses and customers, Voice AI connects us through something even more fundamental — conversation.

“When we first started using AI, it was a tool to remove background noise, improve lighting, make images look better,” said Cullen Jennings, CTO for Audio/Video Collaboration and AI at Cisco. “But it’s shifting now — it’s about how we actually answer our questions, talk to other agents, and communicate with AI itself.”

Cisco’s new Voice AI system operates with latency as low as 1.3 seconds, which Cisco noted is nearly on par with the natural delay between human speakers. That seemingly small number matters — it’s what makes conversations with digital agents feel natural rather than robotic. In early pilots across healthcare and finance, that realism has already led to over 30% higher satisfaction.

Still, the power of Cisco’s AI push isn’t just technical — it’s architectural. By championing open protocols like MCP (Model Context Protocol) and A2A (Agent-to-Agent), Cisco is building a foundation where different AI systems can talk to each other, no matter who made them.

“When Cisco says we’re going to build MCP, we don’t mean it’s only in our product. We mean our products will talk to everybody else’s MCP servers,” Cullen said. “And that’s going to make an ecosystem better than any single company trying to dominate the AI platform. We want to connect all of this together.”

He added that Cisco is contributing these standards to open-source communities to accelerate adoption and ensure interoperability across the broader AI ecosystem — a signal that Cisco sees its future not as a walled garden, but as part of an interconnected web of intelligent systems.

Cisco also showcased how its AI strategy is extending into hardware, embedding intelligence directly into its collaboration devices. The new RoomOS 26, running on Room Bar devices powered by NVIDIA processors, enables on-device inference — reducing cloud dependency, speeding up response times, and improving data privacy.

This hardware-level intelligence gives agents like the Director Agent the ability to instantly frame speakers and manage lighting adjustments, bringing production-quality video to everyday meetings.

“We took a really strong position from the beginning that if we couldn’t do this in a responsible way, it wasn’t worth doing,” Cullen said.

Cisco’s Responsible AI Framework builds on that belief. It’s designed to ensure trustworthiness, transparency, and fairness are embedded in every AI decision — from customer support interactions to workplace automation.

“It’s got to give reasonable answers — answers that are verifiable in a way that a human can check,” Cullen explained. “Bias is inevitable in training data, but you can fight against it. Our AI group spends a lot of time thinking about how to collect and manage data to reduce bias.”

It’s a reminder that for Cisco, the story of AI isn’t just about capability — it’s about accountability.

Most of Cisco’s new AI Agent features, Voice AI capabilities, and the Contact Center AI Suite will begin rolling out to enterprise customers in early 2026, with several components already available in limited beta programs.

To close out WebexOne 2025, acclaimed actor, producer, and entrepreneur Ryan Reynolds brought humor, authenticity, and insight to the main stage. Known for blending creativity with business innovation, Reynolds shared stories from his ventures in film, advertising, and even football—offering a candid look at how curiosity, storytelling, and experimentation drive success across industries.

Reynolds also reflected on the rapid rise of AI in entertainment, joking about how artificial intelligence can now translate his performances into nearly any language—“Tamil, German, French… even French Canadian,” he quipped, earning laughs from the audience. But his tone quickly shifted to something more thoughtful as he considered the broader implications.

“It’s pretty remarkable, but it’s also tricky,” he said. “Like, you get put in a tricky position. Suddenly I go, well, there’s a hundred voice actors here, so what do we do? I always Webex with them all.”

His remarks struck a balance of humor and humility, underscoring how AI collaboration brings both promise and complexity. The moment tied perfectly into Cisco’s theme of Connected Intelligence—a reminder that while technology keeps advancing, human empathy, creativity, and leadership remain at the center of meaningful innovation.

Q: What did Cisco announce at WebexOne 2025?
A: Cisco introduced its Connected Intelligence strategy — a suite of AI agents, analytics tools, and Voice AI systems designed to unite humans and digital workers under one collaborative platform.

Q: How is this different from last year’s focus?
A: “Last year at WebexOne, the theme was ‘Experience Matters,’ which focused on human-to-human collaboration,” said Aruna Ravichandran, CMO of Cisco Collaboration. “This year, we’re taking it to the next step by facilitating human-to-AI and AI-to-AI collaboration.”

Q: What are some of the new AI agents?
A: “We’re launching a lot of agents, not assistants — agents,” Aruna explained. “My top favorites are the Note Taker Agent, Polling Agent, Receptionist Agent, and Director Agent — all built to enhance meeting engagement and collaboration workflows.”

Q: How is Cisco addressing ethics and trust in AI?
A: “If we couldn’t do this responsibly, it wouldn’t be worth doing,” said Cullen Jennings, CTO of Cisco Collaboration. “Long-term, if people lose faith that you’re doing the right thing, nothing else matters. Our Responsible AI Framework focuses on trust, transparency, and bias mitigation.”

Q: What makes Cisco’s AI approach unique?
A: “Our products will talk to everybody else’s MCP (Model Context Protocol) servers,” Cullen explained. “We don’t want to dominate the AI ecosystem — we want to connect it through open standards and interoperability.”

Cisco’s announcements at WebexOne 2025 reflect more than product updates — they mark a philosophical shift toward AI that collaborates, learns, and evolves responsibly.

By focusing on agentic intelligence, open standards, and on-device performance, Cisco is positioning itself as a leader in enterprise AI — one that values connection over control.

As Aruna put it, “It’s no longer just about human-to-human collaboration — it’s people and AI together.” Combined with Cullen’s focus on ethical design and open-source interoperability, Cisco’s direction signals a future where AI systems work with us, not just for us.

Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing, image, and idea-generation support from ChatGPT, an AI assistant. However, the final perspective and editorial choices are solely Alicia Shapiro’s. Special thanks to ChatGPT for assistance with research and editorial support in crafting this article.