argument: Notizie/News - Digital Governance
Source: Atlantic Council
Freelance policy analyst Jay Morgan proposes a joint legislative working group to strengthen US–EU governance of artificial intelligence safety. Morgan recommends establishing the group within the Transatlantic Legislators’ Dialogue, which connects the US House of Representatives and the European Parliament. Morgan argues that lawmakers could advance cooperation even without executive-led negotiations, beginning with narrowly defined cybersecurity measures. The proposed agenda includes common incident terminology and compatible reporting procedures, acknowledging that existing frameworks differ in their triggers and covered entities. Morgan would extend reporting requirements to affected organizations and model developers whose systems contributed to an incident. Morgan also proposes statutory designation of regulatory and law-enforcement contacts for incidents affecting organizations across the two jurisdictions.
Another proposal would coordinate government AI evaluation benchmarks and work toward mutual recognition of safety testing to reduce duplicate assessments. Morgan advocates giving vetted cybersecurity defenders access to more capable versions of frontier models through trusted-access arrangements. For independent evaluators, Morgan proposes accreditation criteria, minimum access to developers’ systems, financial independence and periodic rotation of evaluation teams. Morgan cautions that perceived attempts to export a regulatory regime and a lack of tangible results have undermined previous transatlantic exchanges. Morgan argues that the proposed agenda would require more frequent engagement than the dialogue’s usual twice-yearly meetings. Initial recommendations on incident reporting and agency contacts could, Morgan suggests, support House draft legislation and a European Parliament initiative requesting a Commission legislative proposal.