{"id":33471,"date":"2026-08-21T11:12:34","date_gmt":"2026-08-21T15:12:34","guid":{"rendered":"https:\/\/www.uvm.edu\/it\/kb\/?post_type=ht_kb&p=33471"},"modified":"2026-08-31T15:40:17","modified_gmt":"2026-08-31T19:40:17","slug":"agentic-ai-guidance","status":"publish","type":"ht_kb","link":"https:\/\/www.uvm.edu\/it\/kb\/article\/agentic-ai-guidance\/","title":{"rendered":"Agentic AI Guidance"},"content":{"rendered":" \t\t
\r\n \t\t\tEarly Release Content<\/span> \t\t\t \t\t\t\t

\r\n \t\t\t\t\tThis is an early release of this KB article and content is subject to change. Please refer back to this article occasionally for updated guidance. \t\t\t\t<\/p>\r\n \t\t\t \t\t\t\r\n \t\t<\/div>\r\n \t\t\n

What is Agentic AI?<\/h2>\n

Agentic AI refers to AI tools that can\u00a0act\u00a0on your behalf, not just answer questions\u00a0or generate output. These tools may use natural language instructions, work through multi-step tasks, interact with other applications, call external tools or APIs,\u00a0modify\u00a0files, access data, write code, or make changes in systems where you have access.<\/span><\/p>\n

Unlike\u00a0Generative AI, an AI agent may continue working through a task after your\u00a0initial\u00a0instruction. For example, an agent might organize files, edit a document, update code, interact with online services, or perform actions using your account permissions. Because of this, agentic AI can be\u00a0a\u00a0powerful\u00a0tool.\u00a0It also creates\u00a0additional\u00a0responsibility and risk\u00a0considerations.<\/span><\/p>\n

At UVM, Agentic AI should be used in ways that support innovation, productivity, learning, and research while protecting University data, systems, people, and institutional responsibilities. Agentic tools should be used thoughtfully, with clear limits, appropriate oversight, and careful attention to UVM policies.<\/span><\/p>\n

Why\u00a0Agentic AI\u00a0requires\u00a0extra care<\/span><\/h2>\n

Agentic AI tools often act using your account, your permissions, and your access to systems or data. If you can open a file, access a system, or view certain data, an AI agent running under your account may be able to do the same unless its access is\u00a0specifically\u00a0limited\u00a0through technical controls and instructions.<\/span><\/p>\n

Actions\u00a0taken by an AI agent should be treated as actions taken by you. If an agent moves, changes, shares,\u00a0deletes, summarizes, or sends information, you\u00a0are responsible for\u00a0making sure that action is\u00a0appropriate\u00a0and authorized.<\/span><\/p>\n

Agentic AI may misunderstand your instructions, act in ways you did not\u00a0expect,\u00a0access information you did not intend to include, or interact with systems outside the original task. This is why human oversight, access limits, logging, monitoring, and careful data handling are essential.<\/span><\/p>\n

Security\u00a0Principles for Agentic AI Use<\/span><\/h2>\n

Treat AI agents as acting on your behalf<\/span><\/h3>\n

Assume that an AI agent is acting\u00a0as\u00a0you when it uses your account,\u00a0credentials, files, applications, or system access. Actions taken by the agent should be considered equivalent to actions taken by the person or system account that authorized it.<\/span><\/p>\n

Practical Guidance<\/i><\/strong><\/p>\n