Who Controls What Your Team Shares with AI? | Cloudflare Zero Trust & AI Gateway Live Demo

Most environments have no network-level policy on which AI tools employees can access or what they can share through them. No visibility. No enforcement at the identity layer. In this 30-minute session, Ganesh Gourappan (GlobalDots Senior Solutions Engineer) walks through exactly how to fix that using Cloudflare Zero Trust and Cloudflare AI Gateway. ── What's covered ── 00:00 The problem: shadow AI, prompt exposure, MCP servers 02:00 Cloudflare One as the platform 13:33 DEMO: Shadow IT analytics — see exactly which AI tools your team uses 16:27 DEMO: Gateway rules — block or redirect unsanctioned AI apps 21:24 DEMO: DLP on ChatGPT — block prompts containing PII or credentials 25:05 DEMO: AI Gateway — jailbreak blocking + cost visibility for AI agents 29:06 DEMO: MCP portals — control which MCP servers agents can access 31:00 GlobalDots: who we are ── Key takeaways ── 1. You can block any unsanctioned AI tool in minutes using Cloudflare Gateway, no custom code required 2. DLP rules can stop employees from sending credentials or PII to ChatGPT 3. Cloudflare AI Gateway puts a security and observability layer between your agents and any LLM provider 4. MCP server access can be governed with the same Zero Trust controls you already use for apps ── About GlobalDots ── GlobalDots is a Cloudflare Powered+ Partner and 2025 EMEA Growth Partner of the Year. Want a deeper dive or a 1-on-1 demo? Reach out: globaldots.com