Governance
Guardrails, approval gates, audit trails, the EU AI Act and shadow AI. How to give agents real access without handing over the keys.
Content Guardrails for AI Agents: Inside the Rule Builder
How input/output content guardrails work for AI agents: keyword, regex, PII and written-policy rules, org vs agent scope, and how Oido Studio builds them.
PPWR Compliance Runs on SKU Data Nobody Wants to Collect
PPWR software automates the paperwork. The recycled content, material and weight data behind every SKU is still collected by hand. Here's how an agent fixes that.
Batch Traceability Without an MES: The 4-Hour Test
Your traceability is only as good as your last mock recall. How to trace a lot in four hours with paper and spreadsheets, and where the chain actually breaks.
UK E-Invoicing Mandate: Your First Deadline Isn't 2029
The UK mandate lands April 2029 via Peppol. If you sell into the EU, a nearer date already applies to you. What UK finance teams should do now.
How to Test an AI Agent Before You Trust It With Real Work
A business acceptance test for AI agents, build a golden set from real cases, run it ten times, and pass three gates before the agent touches anything live.
AI Agent Monitoring: How to Know Yours Still Works
Agents fail silently, green dashboards, wrong outcomes. The six signals that tell you an AI agent is still working, and the review that catches drift early.
MCP Server Security: What to Check Before You Connect
Most MCP security guides are written for the team running the server. You are probably the client. The controls that actually reduce blast radius.
Shadow AI Policy: Absorb the Demand, Don't Block It
Blocking shadow AI just moves it to phones. What the 2026 data shows, why per-seat licensing causes it, and a 30-day plan to absorb it into a governed platform.
Why AI Agent Pilots Die Before Production
Most AI agent pilots never ship. The cause is rarely the model, it's the exception path, permissions and audit trail nobody scoped. Here's the fix.
EU AI Act & AI Agents: What Applies From Today
From 2 August 2026 the AI Act is enforceable and Article 50 transparency applies. What that actually means if you run AI agents in your business.
AI Browser Agents: 4 Controls That Actually Work
Letting an agent act inside your logged-in browser is a real risk. The four architectural controls that contain it, and when to use an API instead.
EU E-Invoicing Mandates 2026: Compliant ≠ Automated
Belgium, Poland and France go live in 2026. Compliance is the easy part. What happens to AP once structured invoices arrive is the work.
AI Agent Guardrails: Autonomy Your Auditor Can Live With
The five guardrails every production AI agent needs, scoped tools, permissions, approval gates, budgets, audit logs, and why they create autonomy.
Human-in-the-Loop AI: Full Autonomy Is the Wrong Goal
Why the best AI automations keep a human on the approvals that matter, where the checkpoints go, how they shrink over time, and what never gets automated.