A US company has developed an AI system, CrimeBuster 7909, that collects information about incarcerated individuals that predicts whether someone is likely to commit another crime if released from pri

Question

Options

• ARequire the company to register the tool with the EU database.
• BRequire the application of privacy enhancing technologies.
• CBe subject to approval by the relevant EU authority.
• DBe banned under the EU AI Act.

• A
  2% (1)
• B
  4% (2)
• C
  2% (1)
• D
  92% (46)

Explanation

Under Article 5 of the EU AI Act, AI systems that assess or predict the risk of a natural person committing a criminal offense - including recidivism prediction tools like CrimeBuster 7909 - fall into the "unacceptable risk" category and are explicitly prohibited, making D correct. Option A is wrong because there is no EU registration database for banned systems; registration applies only to certain permitted high-risk systems. Option B is wrong because privacy-enhancing technologies address data protection (GDPR territory) but cannot rescue a system that is flatly prohibited under the AI Act. Option C is wrong because approval by an authority is a mechanism for high-risk AI systems - banned systems cannot be approved at all; the prohibition is absolute.

Memory tip: Picture the EU AI Act as a four-tier pyramid - Unacceptable → High → Limited → Minimal risk. Criminal recidivism scoring sits at the very bottom of that pyramid (unacceptable), so it gets banned, not regulated, registered, or approved. If an AI predicts who will commit a crime, the EU says "no" - full stop.

No community discussion yet for this question.