Designed to provide organisations with centralised visibility and control over autonomous AI tools, AvePoint has launched AgentPulse Command Centre, an AI agent registry.

The platform addresses growing security and cost management challenges as enterprises deploy AI agents across their operations. AgentPulse enables IT and security teams to track which agents are active, identify users, and monitor access to sensitive data.

The launch follows research indicating significant AI-related security risks. According to data cited by AvePoint, 75 per cent of organisations using AI experienced data breaches in the past year. Gartner predicts 40 per cent of agentic AI projects could be abandoned by 2027 due to inadequate risk controls.

Agentic AI systems operate with greater autonomy than traditional chatbots, making independent decisions and taking actions within business systems. This independence creates new security challenges beyond those addressed by legacy IT controls.

Prompt injection attacks represent a particular concern, where malicious actors manipulate AI decision-making processes by embedding harmful instructions in external data sources. Recent research has identified these attacks as systemic challenges affecting AI-powered browser agents across the industry.

AgentPulse shows which AI agents generate the most activity and access large data volumes. Organisations can use this information to identify unnecessary agents driving unexpected costs and adjust or remove them.

The tool also identifies oversharing instances where agents have excessive access to sensitive files. Teams can remediate these issues directly within the Confidence Platform.

"As organisations rapidly deploy AI agents to drive productivity, they're discovering that these powerful tools introduce significant security and data governance challenges that legacy IT controls weren't designed to address," said Jeremy Thake, Chief Architect, AvePoint.

"As part of the Confidence Platform, AgentPulse delivers comprehensive visibility so IT and security teams can confidently scale agentic AI and further innovation, without compromising data security or racking up extra costs."

John Peluso, Chief Technology Officer at AvePoint, noted the financial implications extend beyond licensing costs. Without oversight, organisations face unexpected charges from high-activity agents, storage costs from redundant data, and potential breach expenses.

The solution follows an October update to the Confidence Platform that streamlined compliance enforcement for AI agents. AgentPulse supports multiple agent types including Microsoft 365 Agents Toolkit, Microsoft 365 Copilot Studio and Copilot Studio Lite, SharePoint, Azure AI Foundry, ISV Store, and more.

Gartner analysts have warned about "agent washing," where vendors rebrand existing tools as agentic AI without delivering autonomous capabilities. The analyst firm estimates that only about 130 vendors out of thousands claiming agentic solutions actually offer genuine agentic features.

AvedPoint AgentPulse is available in private preview. Organisations can request access through AvePoint's website.

https://www.avepoint.com/solutions/ai-agent-governance-and-security