On March 14th, 2025, Bússola Tech hosted a panel discussion titled "Usage of Artificial Intelligence in Legislative Drafting: Dos and Don’ts." The session brought together legislative experts and institutional leaders to reflect on how AI can support, challenge, or reshape the delicate craft of legislative drafting within parliamentary institutions.

Moderated by Mr. Wade Ballou, former Chief Legislative Counsel of the U.S. House of Representatives, the discussion featured distinguished speakers: Ms. Donalene Roberts, Legislative Counsel at the U.S. House of Representatives, Mr. Grant Vergottini, CEO of Xcential Legislative Technologies, and Mr. Miguel Landeros Perkic, Secretary-General of the Cámara de Diputadas y Diputados of Chile.

The panel began with each participant outlining their current drafting process and highlighting key operational challenges. From institutional bottlenecks to evolving demands in legislative complexity, the speakers provided first-hand perspectives on the realities of legal text production in their respective jurisdictions.

In the second segment, the discussion pivoted toward an ideal scenario: how each panellist would redesign their drafting systems from the ground up. Their responses underlined the importance of interoperability, clarity in legislative intent, and the need for digital infrastructure that respects institutional values while enabling innovation.

The third round focused on the potential risks of new AI-based systems to the legislative profession and to the broader institution. Concerns raised included the erosion of institutional memory, the opacity of algorithmic decision-making, and the danger of diminishing the human element that safeguards legal accuracy and democratic intent in legislative texts.

Next, the panellists examined how their current processes could be incrementally improved. They considered pragmatic enhancements, such as the integration of AI-assisted tools that complement – rather than replace – legislative drafters, alongside structured frameworks that ensure accountability and transparency in system outputs.

The discussion concluded with a final round in which each speaker shared a key “do” or “don’t” for institutions considering the integration of AI into the drafting process. The closing remarks emphasised the need for cautious optimism, strategic foresight, and a principled approach to innovation that prioritises institutional integrity, legal precision, and the enduring values of democratic governance.

The panel underscored the vital role of collaboration between technologists, legislative staff, and institutional leaders to ensure that any transformation driven by AI remains aligned with the constitutional role and responsibilities of parliaments.