Cisco Systems Inc. said today it’s acquiring the artificial intelligence-native observability startup Galileo Technologies Inc. in order to make “agentic workforces” more trustworthy and reliable.

Galileo is a developer of tools for observing and evaluating AI models, enabling companies to monitor multiagent systems in real time and apply guardrails to ensure the accuracy of their work. The platform has been widely adopted by enterprises over the past couple of years for the way it instills trust in AI agents, and those capabilities will soon be added to Cisco’s Splunk observability platform.

In a blog post, Kamal Hathi, senior vice president and general manager of Splunk, said Galileo was built to solve the problem of trust in AI. “From day one, its platform has given AI teams the tools to evaluate AI quality, detect AI failures before they reach users, and continuously improve AI behavior in production – turning observability from a nice-to-have into a core pillar of AI development,” he said.

According to Hathi, the plan is to use Galileo’s capabilities to enhance Splunk’s observability portfolio. It will result in better AI agent monitoring, with real-time visibility into their actions and more security within the agent development lifecycle.

He explained that Galileo gives developer teams a single platform through which they can instrument every stage of agent development with enterprise-grade rigor. “It is a complete solution that enables deeper insights from the earliest stages of prompt optimization and model selection, through evaluations, all the way to production monitoring, observability and enforcing guardrails,” Hathi added.

The price of the acquisition was not disclosed, but Cisco expects to wrap up the deal by the end of the fourth quarter.

Cisco has collaborated with Galileo in the past. About a year ago, the two companies joined a consortium called AGNTCY, which was set up to create specifications and reference implementations of an open-source architecture for building trustworthy AI agents. AGNTCY will tap Galileo’s security and observation tools and LangChain for its orchestration capabilities.

The networking giant has been trying to position itself as a kind of security layer for AI agents. Last month at RSCA 2026, it unveiled a new Duo Agentic Identity offering that can help enterprises to discover, identify and monitor AI agents being used in their systems. It also announced plans to expand its AI Defense platform to AI agents.

Hathi said AI agent security is an entirely different ballgame than what enterprises are used to. He explained that agentic applications can behave in strange ways, leading to some very unexpected, inaccurate, low-quality and sometimes even harmful outputs.

“These issues can ultimately lead to decreased customer trust, poor end-user experiences and increased costs,” he said. “As a result, teams need visibility across the AI stack beyond signals like latency and errors. Observability must evaluate issues like hallucinations and bias, security metrics to detect and mitigate business risks, and track cost and usage metrics to ensure clear ROI.”

The dangers of AI agents were elaborated on by Cisco President and Chief Product Officer Jeetu Patel during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio, at RSAC last month. Patel likened them to teenagers: “They’re superbly and supremely intelligent, but they have no fear of consequences,” he said. “They do stupid stuff all the time. The difference between delegation and trusted delegation is equivalent to the difference between bankruptcy and market leadership. It is literally that stark.”

Cisco said Galileo’s platform and services will continue to operate independently until the transaction is wound up, but it’s not clear what will happen to existing customers after the deal is finalized.

Image: Galileo