Safe AI Agents: When the AI Can Click Buttons

The AI in your browser, your inbox and your automations no longer just answers, it acts: it fills forms, sends messages, downloads files and updates records while you get on with something else. This course is the safety layer that sits across every agent tool an events team is likely to meet, from Claude Cowork and Claude in Chrome to ChatGPT's Chrome extension, Copilot Autopilot and the AI steps in Zapier, Make and n8n. It covers what an agent can actually reach, how a hidden line in a venue listing or a supplier PDF can hijack it, how to set approvals so a person signs off anything that leaves the building, the four jobs that should never run unsupervised, and a practical checklist for attendee data under UK GDPR (the UK General Data Protection Regulation). It builds on the product courses rather than repeating them, and links to them throughout.

intermediate · 45m · 3 modules

1. What you are handing over

2. Keeping a person in charge

3. Attendee data