EU consultation on high-risk AI classification: blind spot on agentic systems like Pegasus (euobserver.com)

• EU consultation on high-risk AI guidelines closes 23 June 2026
• Guidelines lack distinction between deterministic and agentic AI systems
• Pegasus spyware cited as example of unclassified agentic system
• Proposed remedies: delegated act, standardization mandate, coordination protocol

"On 19 May 2026, the European Commission opened a targeted consultation on draft guidelines for classifying high-risk AI systems under Annex III of the AI Act, closing 23 June. The guidelines fail to distinguish between deterministic workflows and agentic systems that generate steps at runtime. This gap prevented Pegasus spyware from being classified as high-risk. The author proposes using Article 7 delegated acts, a common technical standard mandate, and a coordination protocol to address the oversight."