Source: OJ L, 2024/1689, 12.7.2024

Current language: SV

Artikel 90 Varningar från den vetenskapliga panelen om systemrisker


Summary What does Article 90 of the AI Act say?

This article establishes the mechanism by which the scientific panel can formally flag concerns about general-purpose AI models to the AI Office.

It sits within the broader enforcement section dealing with general-purpose AI models and connects directly to Article 51, which sets out the conditions for classifying a model as presenting systemic risk.

In essence, the scientific panel acts as an early warning mechanism, and a qualified alert from it can trigger the Commission's investigative and enforcement powers laid out in the surrounding articles (Articles 91 to 94).

Important points:

  • The scientific panel is empowered to issue a qualified alert to the AI Office where it suspects a general-purpose AI model poses a concrete identifiable risk at Union level or meets the systemic risk classification conditions of Article 51.
  • Upon receiving a qualified alert, the Commission may exercise its enforcement powers under this Section to assess the matter, and the AI Office must keep the Board informed of any resulting measures.
  • A qualified alert must be duly reasoned and include at minimum the provider's contact details, a description of the relevant facts and reasons for the alert, and any other information the scientific panel considers relevant.

Springlex's summary of the article, a reading aid, not a substitute for the legal text.

    1. Den vetenskapliga panelen får lämna en kvalificerad varning till AI-byrån om den har anledning att misstänka att

      1. en AI-modell för allmänna ändamål utgör en konkret identifierbar risk på unionsnivå, eller

      2. en AI-modell för allmänna ändamål uppfyller de villkor som anges i artikel 51.

    1. Efter en sådan kvalificerad varning får kommissionen, genom AI-byrån och efter att ha informerat styrelsen, utöva de befogenheter som fastställs i detta avsnitt i syfte att bedöma ärendet. AI-byrån ska informera styrelsen om alla åtgärder enligt artiklarna 91–94.

    1. En kvalificerad varning ska vara vederbörligen motiverad och innehålla åtminstone följande uppgifter:

      1. Kontaktpunkten för leverantören av den berörda AI-modellen för allmänna ändamål med systemrisk.

      2. En beskrivning av relevanta fakta och skälen till den vetenskapliga panelens varning.

      3. All övrig information som den vetenskapliga panelen anser vara relevant, inbegripet, i förekommande fall, information som samlats in på eget initiativ.

We're continuously improving our platform to serve you better.

Your feedback matters! Let us know how we can improve.

Found a bug?

Springflod is a Swedish boutique consultancy firm specialising in cyber security within the financial services sector.

We offer professional services concerning information security governance, risk and compliance.

Crafted with ❤️ by Springflod