#16 (EN) IT Leadership News August 2026

In this episode of Me, Myself & IT Leadership, Daniel and Nova discuss the IT leadership news for August 2026 — and the pattern behind a set of very different stories: provenance, responsibility, data flows, and the real cost of AI.

They talk about Anthropic’s invisible watermarking in the context of the EU AI Act, AI agent security at Black Hat, mysterious anonymous models on OpenRouter, Stripe’s acquisition of OpenRouter, Atlassian’s updated training clauses, humanoid traffic police robots in China, and Anthropic’s potential IPO.

The central question of the episode: Do you actually know what you are working with — which model, which provider, which data chain, which cost logic, and which responsibility sits behind it?

Key topics:

- Why an AI watermark does not automatically prove that a text was fully written by AI

- What the EU AI Act changes for AI-generated content

- Why agent security is becoming its own discipline

- The risks created by agent frameworks, skill marketplaces, and AI-generated security patches

- Why free anonymous frontier models can become a serious shadow-IT and data-loss risk

- What Stripe’s acquisition of OpenRouter could mean for AI routing, token costs, and FinOps

- Why vendor change notices need to become part of continuous AI governance

- What IT leaders can practically take away from this month’s news

Take-away: For every AI use case, leaders should be able to answer three questions: Who built it? Where does my data go? And what does it really cost — not per token, but per completed task?