On December 16th, a panel of senior parliamentary officials and legislative technologists from across the Portuguese-speaking world offered a rare, comparative view into how legislatures are digitizing their work—and what breaks when they do. The conversation brought together Luís Kimaid (Bússola Tech) as moderator, Pedro de Neri, Secretary-General of Angola’s National Assembly, Luiz Fernando Bandeira de Mello, former Secretary-General of the Brazilian Federal Senate, Hugo Tavares of Portugal’s Assembly of the Republic, Ambrósio Alves Soares of Mozambique’s Assembly of the Republic, and Juliano Bringer of Ágape Consulting.
Read more: What Congress Can Learn from Lusophone Parliaments About Modernization and AITaken together, their experiences span parliaments at very different stages of institutional maturity—but they surface a common set of second-order lessons that should resonate in Washington.
Across countries, the panelists converged on a counterintuitive insight: technology is rarely the hard part. The real constraints are political sequencing, procedural design, and institutional trust. Angola and Brazil demonstrated that starting with administrative dematerialization—budget workflows, document circulation, signatures—created the political and cultural conditions necessary to later digitize the core legislative process itself. Portugal, by contrast, illustrated a different challenge: once systems mature, frequent changes to rules and procedure become the primary obstacle to further automation.
The discussion also punctured a common assumption—that resistance to modernization is generational. Mozambique’s experience showed that younger members can be among the most skeptical of digital processes, while senior leadership may be the strongest champions, especially when reforms are tied to budget authority and institutional legitimacy.
Finally, the panel offered a sobering lesson on artificial intelligence. Across jurisdictions, legislatures are converging on a cautious, governance-first approach: using AI for transcription, accessibility, and analysis, while insisting on sovereign control over legislative data and decision-support systems.
Executive Summary: Tools for the Modernization of Portuguese-Speaking Parliaments
1. Parliamentary Digitalization as Institutional Transformation
Across Portuguese-speaking parliaments, digitalization is treated as a long-term institutional transformation rather than a technical upgrade. Successful approaches emphasize phased implementation, political leadership, and cultural change.
Examples:
– Angola: Since 2013, administrative processes are ~85% digital and legislative processes 40–50%; the plenary is fully paperless.
– Brazil: The legislative process is nearly fully digital, with electronic signatures, remote voting, and real-time public transparency.
– Portugal: Digital tools are mature and routine, with MPs actively demanding technological solutions.
– Mozambique: Still paper-heavy, but pursuing a leadership-driven mandate for full parliamentary digitalization.
2. Political Buy-In and Change Management
Participants agreed that political buy-in is the hardest challenge. Effective strategies rely on targeting influential MPs, gradual transitions, and leveraging moments of institutional change.
Examples:
– Angola: Focus on party leaders and committee secretaries, using cost savings and international comparisons.
– Brazil: A 15-year transition with system coexistence, accelerated by the COVID-19 pandemic.
– Mozambique: Senior leadership support contrasts with resistance from some younger MPs; budget negotiations are key leverage points.
– Portugal: Low MP resistance, but high complexity from frequent procedural rule changes.
3. Cost, Efficiency, and Sustainability Benefits
Digitalization produces significant cost savings, operational efficiency, and environmental benefits, strengthening the political case for reform.
Examples:
– Angola: Major reductions in paper, toner, logistics, and distribution burdens.
– Brazil: Improved accuracy, version control, hyperlinking of legal references, and reduced transcription errors.
– Mozambique: Heavy printing demands underscore the urgency of transition.
4. Transparency, Public Access, and Democratic Accountability
Modernization increases public visibility and accountability, making transparency both an opportunity and an obligation.
Examples:
– Brazil: Real-time publication of votes, searchable transcripts, and segmented session videos.
– Portugal: Warns that citizens increasingly use external AI tools to interpret parliamentary activity, risking misinterpretation if parliaments do not provide authoritative data.
– Angola & Mozambique: Emphasize that openness to public scrutiny is now unavoidable.
5. Archives, Institutional Memory, and Preservation
Modernization extends to preserving parliamentary history and institutional memory.
Examples:
– Mozambique: Digitizing records back to 1977 to prevent document loss and preserve constitutional and legislative history.
6. Artificial Intelligence: Opportunities and Governance
AI is approached cautiously, with consensus that governance frameworks must precede large-scale deployment.
Examples:
– Portugal: Prioritizes internal committees, clear rules, and controlled data environments; current use focuses on transcription and captioning.
– Brazil: Sees value in transcription, speech preparation, budget analysis, and investigations, but not wholesale legislative drafting.
– Private sector: Advocates hybrid models combining internal interfaces with external AI engines under strict data control.
7. Security, Sovereignty, and Data Control
Cybersecurity concerns strongly shape modernization strategies.
Examples:
– Portugal and Mozambique: Past cyberattacks reinforce preferences for in-house or tightly controlled systems.
– Broad consensus: Commercial AI tools may suit non-sensitive tasks, but core parliamentary systems require sovereign control.
This blogpost is unusual for us, so a word on its methodology. First, we used an AI tool to generate a transcript (in Portuguese) of the conversation. Then we used another AI tool to translate the conversation into English. We then had it generate a lengthy executive summary organized by theme (which we published above). Finally, we had it pull out key points from the executive summary, which we worked into the introduction. Throughout the process, we prompted the AI to focus on what we thought was important. We edited the results and published them here.
Up to this point, everything on this website is the result of our note-taking and analysis. In the past we have used AI to help edit what we write, but it generally does not write anything for us. We are experimenting with co-writing with AI tools as a way of addressing our limited staff capacity. Please let us know what you think.
