Sign in Console

Blog

Research on AI app economics, threat write-ups, and walkthroughs of GATED in simulated environments. Nothing here is presented as a customer's result unless it says so.
4 posts
Case StudySimulated environment5 min read

Simulated Case Study: Securing a Lovable AI App in One Prompt

We put GATED in front of ChefMate AI, the demo recipe app we built on Lovable and Supabase, then sent it prompt theft, scraper and fake coupon requests to see what reached the AI. Nothing did.

Simulated run: every prompt-theft and scraper request blocked in about 0.01 ms, at zero AI cost
October 24, 2026Read
Market Impact6 min read

How Unscreened Chat Traffic Hurts AI App Profit Margins

A straightforward look at AI token pricing and how a small number of oversized messages or off-topic users can wipe out a $19/month subscription.

45 oversized messages can erase a $19/mo subscription
October 22, 2026Read
Threat Research5 min read

How Attackers Trick Storefront Chatbots Into Giving Out Fake Discounts

Examples of how people try to talk retail AI assistants into generating 100%-off promo codes or unauthorized refunds, and how pre-screening stops those messages.

5 common storefront chatbot abuse patterns blocked
October 19, 2026Read
Architecture4 min read

Rule-Based Edge Screening vs. Calling a Second AI Model for Moderation

A practical comparison of speed and cost between running a second AI model on every message versus checking messages at the network edge.

0.008 ms check time vs. 350 to 600 ms for a second AI call
October 15, 2026Read