Skip to content
Spektor

Editorial

Analysis and observations from Spektor on the governance, adoption and critical thinking of artificial intelligence. These texts are written here first, not content pulled in from other channels.

New texts appear here when there is something to say, not on a schedule.

  1. Consent collected for a diagnosis does not authorise training an algorithm

    A class action in the United States accuses Tempus AI of using genetic data collected for diagnosis to train artificial intelligence models and for commercial agreements with the pharmaceutical industry, without new consent. The case concerns any organisation considering reusing data collected for one purpose.

    Raffaella Aghemo

  2. Who decides whether a system is "high-risk" does not just look at the code

    The European Commission has published draft guidelines on classifying high-risk AI systems. The most relevant point is not technical: it concerns how an organisation describes its systems, wherever it does so.

    Raffaella Aghemo