Most of the public conversation about AI over the last two years has been about generative systems: a chatbot that drafts an email, summarises a report, answers a question when you put one to it. That conversation is already behind where the technology has moved. The frontier has shifted to agentic AI: systems built to take a goal, break it into steps, and execute those steps across real software environments with minimal human involvement in between.

Anthropic announced that Claude will start watermarking the text it generates. This means an invisible, machine readable mark is woven into the text itself. It applies to every Claude model launched on or after 2 August 2026, and it applies globally rather than only to users in Europe.

What happens when AI gets it wrong? Who is affected, and how can it be fixed?

In our last post we established that AI is not one thing. It is a spectrum of technologies, from the algorithm that filters your spam to systems that can analyse medical scans and make decisions that affect people's lives.

For those of you who have just started to engage with the world of AI or you are exploring it for the first time, we wanted to start with a piece that sets out the basics.