AI Safety Policy Must Train Legal Practitioners to Bridge Theory-Practice Gap
A law grad exposes why GDPR-style implementation failures are repeating in AI policy.
Katalina Hernandez, a London law graduate, recounts how her legal education repeatedly prioritized exam-ready theory over real-world practice. Barristers who taught from actual case files were reprimanded for deviating from the curriculum. This systemic disconnect, she argues, is now dangerously replicating itself in AI safety policy. The field needs professionals who are academically literate enough to read cutting-edge AI research, yet close enough to practice to anticipate how rules will be circumvented. Currently, the few hybrids with legal and technical skills gravitate toward high-level policymaking or standards writing, leaving a critical gap on the implementation side—the people who will hear AI liability cases or approve AI agent deployments.
Hernandez uses the GDPR’s Data Protection Officer (DPO) mandate as a cautionary tale. On paper, Articles 37-39 required independent, high-reporting DPOs to enforce compliance—a policy triumph. In practice, enforcement was inconsistent, and many DPOs lacked the authority or resources to push back against business interests. The lesson: even a well-designed policy fails if the implementers aren’t trained and empowered. AI safety risks repeating this if we don’t invest in training legal practitioners who understand both the technical substratum (e.g., multi-agent systems, model evaluation frameworks) and the operational realities of litigation, licensing, and market oversight.
The author calls for a deliberate pipeline: legal education reform, practical training modules on AI systems, and career pathways that place technically literate lawyers in courts, regulatory bodies, and corporate compliance roles—not just in think tanks or policy salons. Without bridging the theory-practice divide, the upcoming wave of AI-related cases and enforcement actions will amplify the same systemic failures seen in data protection.
- Katalina Hernandez highlights that legal curricula often ignore real-world practice, creating a dangerous gap for AI safety implementation.
- GDPR's Data Protection Officer mandate succeeded in law but faltered in enforcement due to lack of skilled, independent practitioners—a pattern AI policy risks repeating.
- The field needs hybrid professionals—technically literate lawyers who can move between policy rooms and courtrooms—but current incentives push them toward policy writing, not practice.
Why It Matters
Without trained legal practitioners, AI safety policies will be circumvented or unenforced—repeating GDPR’s implementation failures at much higher stakes.