The dialog round AI and its enterprise functions has quickly shifted focus to AI brokers—autonomous AI programs that aren’t solely able to conversing, but additionally reasoning, planning, and executing autonomous actions.
Our Cisco AI Readiness Index 2025 underscores this pleasure, as 83% of corporations surveyed already intend to develop or deploy AI brokers throughout quite a lot of use circumstances. On the identical time, these companies are clear about their sensible challenges: infrastructure limitations, workforce planning gaps, and naturally, safety.
At a cut-off date the place many safety groups are nonetheless contending with AI safety at a excessive stage, brokers develop the AI danger floor even additional. In any case, a chatbot can say one thing dangerous, however an AI agent can do one thing dangerous.
We launched Cisco AI Protection at first of this yr as our reply to AI danger—a very complete safety resolution for the event and deployment of enterprise AI functions. As this danger floor grows, we need to spotlight how AI Protection has advanced to satisfy these challenges head-on with AI provide chain scanning and purpose-built runtime protections for AI brokers.
Under, we’ll share actual examples of AI provide chain and agent vulnerabilities, unpack their potential implications for enterprise functions, and share how AI Protection permits companies to straight mitigate these dangers.
Figuring out vulnerabilities in your AI provide chain
Fashionable AI improvement depends on a myriad of third-party and open-source elements comparable to fashions and datasets. With the appearance of AI brokers, that listing has grown to incorporate belongings like MCP servers, instruments, and extra.
Whereas they make AI improvement extra accessible and environment friendly than ever, third-party AI belongings introduce danger. A compromised element within the provide chain successfully undermines your complete system, creating alternatives for code execution, delicate information exfiltration, and different insecure outcomes.
This isn’t simply theoretical, both. A couple of months in the past, researchers at Koi Safety recognized the primary recognized malicious MCP server within the wild. This bundle, which had already garnered hundreds of downloads, included malicious code to discreetly BCC an unsanctioned third-party on each single e mail. Related malicious inclusions have been present in open-source fashions, device recordsdata, and numerous different AI belongings.
Cisco AI Protection will straight tackle AI provide chain danger by scanning mannequin recordsdata and MCP servers in enterprise repositories to determine and flag potential vulnerabilities.
By surfacing potential points like mannequin manipulation, arbitrary code execution, information exfiltration, and gear compromise, our resolution helps stop AI builders from constructing with insecure elements. By integrating provide chain scanning tightly throughout the improvement lifecycle, companies can construct and deploy AI functions on a dependable and safe basis.
Safeguarding AI brokers with purpose-built protections
A manufacturing AI software is prone to any variety of explicitly malicious assaults or unintentionally dangerous outcomes—immediate injections, information leakage, toxicity, denial of service, and extra.
After we launched Cisco AI Protection, our runtime safety guardrails had been particularly designed to guard in opposition to these eventualities. Bi-directional inspection and filtering prevented dangerous content material from each consumer prompts and mannequin responses, retaining interactions with enterprise AI functions secure and safe.
With agentic AI and the introduction of multi-agent programs, there are new vectors to contemplate: better entry to delicate information, autonomous decision-making, and sophisticated interactions between human customers, brokers, and instruments.
To satisfy this rising danger, Cisco AI Protection has advanced with purpose-built runtime safety for brokers. AI Protection will operate as a form of MCP gateway, intercepting calls between an agent and MCP server to fight new threats like device compromise.
Let’s drill into an instance to higher perceive it. Think about a device which brokers leverage to look and summarize content material on the internet. One of many web sites searched comprises discreet directions to hijack the AI, a well-recognized situation generally known as an “oblique immediate injection.”
With easy AI chatbots, oblique immediate injections would possibly unfold misinformation, elicit a dangerous response, or distribute a phishing hyperlink. With brokers, the potential grows—the immediate would possibly instruct the AI to steal delicate information, distribute malicious emails, or hijack a linked device.
Cisco AI Protection will defend these agentic interactions on two fronts. Our beforehand present AI guardrails will monitor interactions between the applying and mannequin, simply as they’ve since day one. Our new, purpose-built agentic guardrails will study interactions between the mannequin and MCP server to make sure that these too are secure and safe.
Our objective with these new capabilities is unchanged—we need to allow companies to deploy and innovate with AI confidently and with out worry. Cisco stays on the forefront of AI safety analysis, collaborating with AI requirements our bodies, main enterprises, and even partnering with Hugging Face to scan each public file uploaded to the world’s largest AI repository. Combining this experience with a long time of Cisco’s networking management, AI Protection delivers an AI safety resolution that’s complete and performed at a community stage.
For these excited by MCP safety, try an open-source model of our MCP Scanner which you can get began with right now. Enterprises on the lookout for a extra complete resolution to deal with their AI and agentic safety considerations ought to schedule time with an professional from our workforce.
Most of the merchandise and options described herein stay in various levels of improvement and will probably be supplied on a when-and-if-available foundation.
