The Finance edition
The AI news of that week, in plain language.
This week is about rules, risks, and sums: the EU law starts enforcing, scammers get smarter, and AI creeps deeper into Excel and your bookkeeping.
EU AI Act starts enforcing: from 2 August you must know which AI you use
The first rules of the European AI Act are now really being checked, and that affects accounting firms too.
From 2 August 2026, the Dutch regulators (AP and RDI) will start enforcing the first rules of the EU AI Act. You will then need to keep track of which AI systems you use and for what. If you use a chatbot, it must be clear that it is AI (the transparency duty in Article 50). Article 4 also takes effect: staff must know enough about AI to use it sensibly. The heavier rules for high-risk AI, such as software that assesses job applicants or credit requests, have been postponed to 2 December 2027.
Background for subscribersPut on a single sheet of A4 which AI tools your office uses and for what purpose. Think of a chatbot on the site or the scan function in your bookkeeping package. That overview is exactly what the law now asks for, and it is handy right away to see where you run risk. Does a client use a chatbot of yours? Make sure it clearly says it is AI. This is not a big project, but you do need it in order from 2 August.
Deepfake fraud grows: fake CEOs and false damage claims
In 2026, scammers increasingly use fake videos and cloned voices to pose as a director or manager. This way they try to push staff into a fraudulent payment. Deepfakes are also turning up in insurance claims and identity checks. Experts advise a 'zero trust' approach: always check every payment request through a second channel agreed in advance.
Background for subscribersThis is a taster.
Subscribers read the whole Finance edition: every story, each with its background in plain language.
Subscribe →