Editorial illustration for 85% of firms run AI agents; 5% trust them to ship, Cisco adds zero‑trust limits
AI Agents: 85% Deploy, Only 5% Trust Production Launch
85% of firms run AI agents; 5% trust them to ship, Cisco adds zero‑trust limits
Why are so many companies still hesitant to let AI agents go live? A fresh survey shows 85 % of enterprises have already deployed agents, yet only 5 % feel comfortable shipping them into production. The gap hints at lingering doubts about control, accountability and the potential for misuse.
While the technology can automate routine tasks, managers worry about agents slipping beyond their intended scope or exposing sensitive data. That unease is prompting vendors to tighten the guardrails around these digital workers. Security teams are looking for ways to grant just‑enough access, enforce time limits and keep a clear audit trail.
In that climate, new offerings that blend identity management with continuous risk assessment are gaining attention. The next step for many firms will be to see whether these controls can bridge the trust deficit and make the agentic workforce a reliable part of everyday operations.
Cisco also extended zero trust to the agentic workforce through new Duo IAM and Secure Access capabilities, giving every agent time-bound, task-specific permissions. On the SOC side, Splunk announced Exposure Analytics for continuous risk scoring, Detection Studio for streamlined detection engineering, and Federated Search for investigating across distributed data environments. The zero-human-code engineering mandate AI Defense, the product Cisco launched a year before RSAC 2026, is now 100% built with AI.
By the end of 2026, half a dozen Cisco products will reach the same milestone. By the end of calendar year 2027, Patel's goal is 70% of Cisco's products built entirely by AI. "Just process that for a second and go: a $60 billion company is gonna have 70% of the products that are gonna have no human lines of code," Patel told VentureBeat.
"The concept of a legacy company no longer exists." He connected that mandate to a cultural shift inside the engineering organization. "There's gonna be two kinds of people: ones that code with AI and ones that don't work at Cisco," Patel said. "Changing 30,000 people to change the way that they work at the very core of what they do in engineering cannot happen if you just make it a democratic process.
It has to be something that's driven from the top down." Five moats for the agentic era, and what CISOs can verify today Patel laid out five strategic advantages that will separate winning enterprises from failing ones. VentureBeat mapped each moat against actions security teams can begin verifying today. The telemetry layer the industry is still building Patel's framework operates at the identity and policy layer.
Are enterprises ready for autonomous code? Eighty‑five percent of firms say they are, running AI‑agent pilots across their stacks. Yet only five percent have let those agents ship, citing trust as the decisive factor.
Cisco’s Jeetu Patel linked that trust gap directly to market survival, warning that without it firms risk bankruptcy. To address the concern, Cisco extended its zero‑trust model to the agentic workforce, rolling out Duo IAM and Secure Access features that grant each agent time‑bound, task‑specific permissions. The move also triggers a sweeping mandate that will reshape Cisco’s 90,000‑person engineering organization.
On the security operations side, Splunk introduced Exposure Analytics for continuous risk scoring and a Detection Studio to streamline detection engineering. Whether these controls will close the trust gap remains uncertain; the technology exists, but adoption lags. Companies must weigh the promise of automation against the practical need they've come to expect for verifiable safeguards.
In short, the numbers reveal enthusiasm, but confidence in production use is still limited.
Further Reading
Related Reading
Business & Startups OpenAI, a Series F San Francisco startup founded in 2015 by eight pioneers
Business & Startups Terminal-Bench 2.0 launches with Harbor, testing any container-installable agent
Business & Startups Zuckerberg Unveils Meta Compute to Build Global AI Infrastructure
From Archives Microsoft adds ’vibe working’ to Word and Excel; Copilot Agent Mode now default
From Archives Industry Shifts to Richer Context for AI Agents, Guided by Human Judgment
Common Questions Answered
Why are only 5% of enterprises comfortable shipping AI agents into production?
Despite 85% of firms deploying AI agents, only 5% trust them for full production use due to concerns about control and potential misuse. Managers worry about agents potentially expanding beyond their intended scope or accidentally exposing sensitive company data.
How is Cisco addressing trust issues with AI agents in enterprise environments?
Cisco extended its zero-trust model to the agentic workforce by introducing Duo IAM and Secure Access capabilities that provide time-bound, task-specific permissions to AI agents. These new features aim to give enterprises more granular control and increase confidence in AI agent deployments.
What are the key challenges preventing widespread AI agent adoption in enterprises?
The primary challenges include concerns about accountability, potential data exposure, and the risk of AI agents operating beyond their intended parameters. Enterprises are hesitant to fully trust AI agents despite recognizing their potential to automate routine tasks.