Beyond the Algorithm: From Research to Action for Democratic Renewal [Slides]

Welcome and Opening Remarks Jennifer Edmond -Trinity College Dublin | KT4D Ilaria Mariani -Politecnico di Milano | ORBIS Welcome! Who is in the room and online Over 200 people registered from over 40 countries around the world Researchers, Civil Society Organisations, Policymakers, Industry, Citizens Basic information for those of you with us here in Brussels Fire exits and toilets 2 Basic information for those of you with us online 04/12/2025 3 The event will be recorded and will be made available afterwards (including the presentations of each speaker) We do encourage you to enter any question in the dedicated Q/A box placed in the lower toolbar. Use the chat to introduce yourself and share your experience and comments. You can follow the chat to be informed and receive link on the Task Force projects’ main outputs and publications. 4 Who is behind the event *AI4Gov - Trusted AI for Transparent Public Governance fostering Democratic Values introduces solutions and frameworks to increase trust in democratic processes and provide policymakers with automated and evidence-based decision-making tools EASSH - European Alliance for Social Sciences and Humanities *ITHACA - artificial Intelligence To enHAnce Civic pArticipation develops a human-centric AI and ethics-by-design online discussion platform for civic participation in local governance. *KT4D - Knowledge Technologies for Democracy explores how knowledge technologies might foster more inclusive civic participation in democracy and how to protect individuals and institutions from their misuse. *ORBIS - Augmenting participation, co-creation, trust and transparency in Deliberative Democracy at all scales shapes and supports new democratic models developed through deliberative democracy processes using advanced AI tools. 02/12/2025 5 The AIBD&D TF and the Democracy Shield Why the Task force was born and its relevance Ensure that the role of AI as an information and opinion intermediary is well understood and adequately guided; Foster greater citizen agency and democratic alignment, fit for purpose for a technologised world, informing policies and regulation measures, building stronger institutions, enhancing capacity in CSOs; and driving human-centric technology development; Develop a nuanced understanding of how democracy responds to technological threats and opportunities, from FIMI to virtual micropublics; Making direct contributions to both the Knowledge Hub and the Stakeholder Platform. 6 Structure of the Day 3 panels organised loosely around key themes representing points of tension and opportunity for AI, big data and democracy Trust, Literacies and Democratic Technologies Broad interpretation and diverse voices, from the projects and beyond Highlighting challenges and solutions, knowledge creation and innovation, as well as future needs: from theory to practice, from methodology to technology! Plenty of time in each panel for discussion and exchange with the panellists A brief introduction to some of the continuing projects Exhibition tables during the breaks and at the end of the day to facilitate deeper exchanges with the projects and their results 7 So let’s begin! 8 9 How ITHACA Helps People Feel Heard and Included •Inclusive & Accessible Design •WCAG-aligned, screen readers, Text-to-Speech, Speech-to-Text •Support for cognitive, linguistic, and digital literacy needs •Respectful & Safe Environment •Toxicity detection for text & images •Warnings, filtering & flagging •Making Sense of Complex Debate •AI summaries, clustering, sentiment analysis •Helps citizens and officials navigate long discussions •Multilingual Engagement & Gamification •Cross-language dialogue •Motivation & recognition for constructive participation Trust forms when participation is accessible, safe, and meaningful. 04/12/2025 16 Losing Trust: Risks and How ITHACA Mitigates Them •Risk of biased or inconsistent moderation •Mitigated by ITHACA’s Fairness Tool: audits models for disparate impact, over-censorship, and unequal treatment. •Risk of privacy breaches •Mitigated through Differential Privacy & privacy-preserving ML. •Risk of opaque AI behaviour •Mitigated through transparent user-facing explanations (when/why content is flagged, summarised, or clustered). •Risk of security vulnerabilities in AI models •Mitigated via ModelScan for malicious code & robustness issues. Trust is lost when people feel unprotected or misrepresented —ITHACA addresses this proactively. 04/12/2025 17 Rebuilding Trust: Governance & Accountability •Citizen Juries on AI •Residents assess risks, impacts & expectations; co-create mitigation strategies. •Participatory foresight & co-design •Ensures AI features align with community values. •Data Governance Framework •Guidance for secure, ethical civic-data handling under EU values (AI Act, DSA, ethics guidelines). •Policy recommendations •Showing how future civic platforms can operationalise trust, fairness & inclusiveness. Trust can be rebuilt through shared oversight and transparent rules —not technology alone. 04/12/2025 18 Key Features Showcased in the Platform ●AI Summarisation of Discussions ●Sentiment & Stance Detection ●Argument Assistance & Suggestions ●Toxicity filter ●Explainability Panels ●Topic Grouping & Insight Generation ●Gamification elements ●AI Chatbot 04/12/2025 19 What ITHACA teaches Us about Trust 1. Trustworthy civic AI = tech + governance + people. Not just accuracy —but fairness, transparency, and shared control. 2. Protection enables participation. Accessibility, safety & privacy are not constraints —they are preconditions for engagement. 3. Regulation needs real-world testing. ITHACA shows how EU principles can be implemented and evaluated in real communities. If we want citizens to trust AI in democracy, we must build AI that earns that trust —visibly, audibly, and continuously. ITHACA is our contribution to that effort. 04/12/2025 20 Artificial Intelligence for Institutionalised, Multimodal, Gamified, Mass Democratic Deliberations 04/12/2025 21 https://www.ai4dproject.eu/ Similar Initiatives What it is: •A Horizon Europe project advancing large-scale democratic deliberation supported by AI. •Focused on multi-modal participation (text, audio, video) and gamified engagement. What it develops: •An AI-enhanced deliberation platform for summarisation, moderation, argument extraction, and safe discourse. •A framework for governments to institutionalise digital deliberation in policymaking. •Real-life pilots on major societal challenges (e.g., climate, public health). Why it matters: •Strengthens democratic trust by making deliberation inclusive, transparent, scalable, and engaging. •Complements ITHACA by exploring AI-supported deliberation at larger scales and institutional levels. 04/12/2025 22 Thank you Epameinondas Koutavelis, Research Project Delivery Manager, Coordinator of ITHACA Konnektable Technologies, LTD. Trusted AI for Transparent Public Governance fostering Democratic Values Dr. Spiros Borotis Senior Project Manager AI4Gov: Trusted AI for Transparent Public Governance fostering Democratic Values December 4, 2025 Democracy Beyond Algorithm 25 1 / 2023 –12 / 2025 Policy recommendations Recommendation 1: Establish Mandatory EU Certification for Bias Mitigation in Public Sector Recommendation 2: Support Capacity Building in Public Administration for Ethical AI Deployment Recommendation 3: Enforce Context-Aware and User-Centric Explainability for AI Decisions December 4, 2025 Democracy Beyond Algorithm 32 Thank you! December 4, 2025 Democracy Beyond Algorithm 33 Website: https://ai4gov-project.eu LinkedIn: https://www.linkedin.com/company/ai4gov-project/ AI4Gov platform: https://cluster-ai4gov.euprojects.net/ Online learning: https://ai4gov-project.eu/home/resources/training-learning/ EU research support to democratic resilience and trust in the digital age “Beyond the Algorithm: From Research to Action for Democratic Renewal” Dr. Anne HAGLUND-MORRISSEY Deputy Head of Unit ‘Democracy, Equality and Culture’ DG Research & Innovation European Commission Policy Context –Current political guidelines and priorities European Union priorities 2024-2029: •A free and democratic Europe (including upholding European values within the EU, and living up to living up to EU values at global level) •A strong and secure Europe •A prosperous and competitive Europe (including bolstering the EU’s competitiveness) European Commission priorities for 2024-2029 include: •Protecting our democracy, upholding our values: Protecting and defending democracy and increasing societal resilience and preparedness,strengthening the rule of law for a fair and wellfunctioning society, championing civic engagement and participation to bring citizens’ ideas to the heart of policy-making Political Guidelines for the European Commission 2024-2029 “Our democratic systems and institutions are under attack. We have seen a rise in the number of threats from internal and foreign actors –whether hostile governments or non-state actors. The methods used are now harder to track, more damaging and easier to deploy with digital tools and social media. This reflects a deep change in the information space, shifting from editorial media sources to user-generated content mediated by platforms and pushed by algorithms. This enables new freedoms but also lowers the cost of manipulating information”. European Democracy Shield –Commission Communication Objective: Boost capacity to counter information manipulation and disinformation and strengthen resilience through a whole-of-society approach. Actions across three main pillars: 1) safeguarding the integrity of the information space; 2) strengthening institutions, fair and free elections, and free and independent media; and 3) boosting societal resilience and citizens' engagement. A new European Centre for Democratic Resilience to bring together EU and Member States' expertise and resources to increase collective capacity to anticipate, detect and respond to threats and build democratic resilience, with a focus on foreign information manipulation and interference (FIMI) and disinformation. AStakeholder Platform will be established within the Centre to facilitate dialogue with trusted stakeholders such as civil society organisations, researchers and academia, factcheckers and media providers Democratic trust, AI and Big Data –Challenging context ▪Big data and AI are shaping our societies at an unprecedented rate: ▪An ever-increasing amount of data is produced revealing people’s attitudes, preferences, views and opinions. Public and private actors collect it and use it in multiple ways: e.g. companies “privatise”data to augment commercial returns; state actors can use it for safety and security applications; and the public sector to provide better, tailored services to citizens. ▪AI and big data open great opportunities in many fields of public interest: education, training, health, safety and security, public services, as well as for democratic processes and civic participation. ▪However, both private and public uses contain some risks at the expense of citizens’rights. These technologies, being at the forefront of datafication processes, pose new challenges both to core individual values such as privacy, freedom and equality, as well as to European collective values, such as fairness, security, inclusiveness, accountability and democratic control. Risks include: uses of AI by non-democratic actors/regimes as a tool of mass manipulation, to spread disinformation, weaken trust, and polarise societies; algorithms can amplify divisive content; bots and fake accounts can drown out real debate (ex. Moldovan 2024 elections showed this clearly: AI-generated content, fake media sites, and cyberattacks targeted democratic integrity). European Democracy Shield - AI and Big Data EDS aims to ensure that AI and big data foster civic participation, transparency, trust and efficiency in democratic governance through a combination of regulatory measures, support for innovation and citizen empowerment initiatives. Through these actions, the EDS seeks to create a robust digital ecosystem where AI and big data serve to strengthen the democratic process, trust and institutions, rather than undermining them: ▪Fostering civic participation: Stimulating civic tech innovation; Enhancing Eparticipation platforms; Promoting digital inclusion; Boosting digital and media literacy ▪Enhancing transparency: Enforcing the Digital Services Act (DSA) and AI Act; Mandating AI-generated content identification; Improving transparency of political advertising; Supporting independent fact-finding and media ▪Improving efficiency in democratic governance: Enabling evidence-based policymaking; Establishing the European Centre for Democratic Resilience; Developing guidance for responsible AI-use CLUSTER 2 CULTURE, CREATIVITY & INCLUSIVE SOCIETY Destination: Innovative Research on Democracy & Governance The research funded in this destination will support the EU ongoing efforts to address threats to democratic systems, and to rebuild citizens’ trust in democracy, including by responding to their social and economic needs. Accelerated and intensifying movement of rejection and challenge of democracy: ●Increased fragility and vulnerability over the last years →number of countries moving towards authoritarianism and overall decline of the quality of democracies. ●Shift from governance based on expertise, rules-based multilateralism and consensual policymaking towards majoritarianism, unilateralism, nationalism, populism and polarisation. ●Levels of trust in the political institutions of democracy decline. ●Increasing spread of misinformation, disinformation and misleading information. Democratic trust, AI and Big Data –Horizon Europe funding Horizon Europe funds AI and Big data in democratic governance through Cluster 2 (Culture, Creativity and Inclusive Society), which specifically allocates funding for calls like “Artificial Intelligence, Big Data and Democracy” (2022), or “Digital Democracy”(2024). Such targeted calls support projects aiming to enhance civic participation, transparency, trust, and efficiency in democratic processes while addressing associated risks. •Civic Participation & Engagement: Projects explore ways to enhance and expand the implementation of civic participation using digital tools, aiming to include less engaged communities and improve the articulation between deliberative processes and representative institutions. •Transparency & Accountability: A major focus is on ensuring transparent and inclusive decision-making processes and addressing potential biases or unfairness in automated decision-making systems. This includes developing methods for constructing meaningful explanations of opaque AI/machine learning systems to empower citizens with the "right of explanation". •Efficiency & Policy Support: Funded research aims to stimulate the effectiveness of public policy by providing policymakers and citizens with tools for evidence-based policymaking, while also taking into account vulnerable groups and cultural aspects. Open Discussion + Q&A For those who join us online, please use the Q&A box to pose your questions 48 Coffee break 10:45 – 11:15 CET 50 Building Resilience: Critical Digital Literacy as Democratic Defence Dr Eleonora Lima (KT4D, Trinity College Dublin) Prof Jennifer Edmond (KT4D, Trinity College Dublin) 4 December 2025, Brussels Democracy Beyond the Algorithm Technical skills are not enough to identify the mismatch between AI tools and cultural values ‘Cultural lag’ (Ogburn, 1922): gap between introducing a piece of technology into society and society’s drawing up rules to manage it Evident mismatch between technology and culture: The Romanian keyboard 53 Apple’s Romanian keyboard with the wrong S (Turkish, not Romanian) What happens when this lack of regard for cultural identities and values is hidden behind a seamless, opaque interface? When a user faces a seamless, opaque interface, their cultural "protective layer" is disrupted They then risk becoming passive consumers rather than active citizens. How do we fix this? We need Critical Digital Literacy We need Critical Digital Literacy From asking: How do I use this? To asking: Who does this serve? Does this align with my community’s values? Is this fit for democracy? IMAGE: Hanna Barakat & Archival Images of AI + AIxDESIGN / https://betterimagesofai.org / https://creativecommons.org/licenses/by/4.0/ We created 3 tools to support Critical Digital Literacy: •For the general citizen who may not have strong technical backgrounds: an immersive escape-room style game. Players take on the role of national leaders facing ethical dilemmas on AI and governance •For citizens who are already comfortable with technology but want to understand more, two interactive explainers on deepfake videos and on recommendation algorithms •For software developers, an interactive red-teaming game to help them see their work not just as code, but as cultural interventions with real societal consequences Ilaria Mariani & Anna Moro, Politecnico di Milano, ORBIS LITERACY Plurality, multitude, diversity, beyond tech focus. 64 LITERACIES Ilaria Mariani & Anna Moro, Politecnico di Milano, ORBIS access participation settings meaningfully participate interpret results understand how AI is in the loop and recognise multi-perspective knowledge construction Literacy is the capacity to access, interpret, adopt and adapt, and thus act within socio-technical systems 65 Ilaria Mariani & Anna Moro, Politecnico di Milano, ORBIS 66 to the ability to access democratic spaces on equal and meaningful terms Digital literacy as access. Entering the space of deliberation from the ability to use a tech system Ilaria Mariani & Anna Moro, Politecnico di Milano, ORBIS where/when access is (still) structurally unequal support ability to access, understand, meaningfully participate 67 Participation literacy. Enabling civic voice in complex debates Ilaria Mariani & Anna Moro, Politecnico di Milano, ORBIS what the discussion is producing where consensus or disagreement is forming what themes are emerging and what patterns matter recognise tensions and minority viewpoints understand why a topic is important 68 Interpretatio n literacy. Making collective reasoning understandabl e Ilaria Mariani & Anna Moro, Politecnico di Milano, ORBIS visual + narrative explainability 69 Explainability literacy. HumanCentred XAI for democratic accountability how/which contributions are grouped opportunities to contest AI interpretations representations of uncertainty and ambiguity I can see how individual voices are transformed into shared knowledge—and I can judge whether that transformation is valid and trustworthy, seeing why the model gets to that output Ilaria Mariani & Anna Moro, Politecnico di Milano, ORBIS participants see summaries, maps, and thematic clusters facilitators see dashboards, tension points, evolving themes 70 Meta-literacy. Understanding who sees what and why policymakers see aggregated insights and actionable evidence Ilaria Mariani & Anna Moro, Politecnico di Milano, ORBIS 71 LEARNING/DELIBERATIVE JOURNEYS through ORBIS digital technologies REIMAGINING THE BUILDING BLOCKS OF DEMOCRACY TEENS FOR DEMOCRACY YOUNG THINKERS AND EU DECISIONMAKERS FUTURE4CITIZENS TRANSITION MECHANISMS FOR A DEMOCRATIC CITY ALLIANCE FOR ENTREPRENEURSHIP AND DEVELOPMENT IN WESTERN GREECE Org. Democracy and Culture Foundation Democracy and Culture Foundation Center for European Policy Studies Re-Imagine Europa Rob De Matt Perifereia Dytiki Ellada Scale International Initially European, to scale out to a global level European Initially national, to scale out to a European level Initially local, to scale out to national and EU level Initially regional, to scale out to the national level Objectiv e Proposing policy recommendations to enhance democratic resilience, initiating global debates via DCF community and Athens Democracy Forum. Engages teenagers (aged 13-17) in policy design through municipal councils in five cities. Aims to bridge the intergenerational gap. Provide a platform for young minds to express innovative ideas and perspectives on the policy scene. Enhance dialogue between citizens and policymakers to foster inclusive, transparent, and trustful deliberative democracy in Europe. Collect diverse voices in Milan to foster democratic processes and participation in urban planning and policy-making. Acts as a network to facilitate the principles of the Small Business Act, promoting economic and social development. Team 2 persons from pilot 2 persons from pilot 4 to 5 persons from pilot 4 persons from pilot 3 persons from pilot 2 persons from pilot Metho d Employs citizen assemblies, structured workshops, and policy advocacy sessions. Integrates top-down and bottom-up approaches.. Annual Teens’ Athens Democracy Forum, preparatory meetings, and collaborative sessions with policymakers. Workshops, online webinars, in-person sessions, and Ideas Lab sessions for public events. Listen-Reframe-Act methodology integrated with ORBIS platform functionalities. Structured dialogues, codesign workshops, and public consultations facilitated by RDM in collaboration with associations. Regional policy recommendation development, collaborative networks of entrepreneurs, and thematic assemblies. P 1 P 3 P 5 P 2 P 4 P 6 WHAT MAKE US FEEL SAFE? A Collective Dialogue on Security in Changing Times Common Pilot International and multiscale Test a multidimensional deliberation on line process on safety perception and collection of ideas 1 persons from each pilot and tech parterns Voting and collection pf input, four online Discussions, online video voting and final policy round table. + 1 Ilaria Mariani & Anna Moro, Politecnico di Milano, ORBIS 72 6+1 Tailored Journeys context-based adaptation plural deliberative pathways non-standardised practices deliberative process and scaling dynamics deliberative steps from user perspective Ilaria Mariani & Anna Moro, Politecnico di Milano, ORBIS 73 Heterogeneous communities Multiple forms of engagement manca foto hybrid participation formats form small focus group to larger online deliberation co-designed experiments fostered engagement and capacitation participants input from voting to collective reshaping of deliberation pioneering contexts to engage those who rarely participate pioneering contexts to engage those who rarely participate Special Eurobarometer: Protecting and promoting democracy 80 Eurobarometer 568 Actions / resources that best equip citizens to address disinformation and information manipulation Eurobarometer: Social Media Survey 2025 81 Flash Eurobarometer FL014EP (October 2025) Eurobarometer: Social Media Survey 2025 82 Flash Eurobarometer FL014EP (October 2025) Latest trends in news sources Main source for news among young people (18-24) is social media & video networks (44%) ahead of News websites/apps (22%) For 45+ News websites/apps remain the main news source. 83 Source: Reuters 2025 Digital News Report Latest trends in news sources Audiences remain sceptical about the use of AI in the news: they expect that the news could become less transparent (-8), less accurate (-8) and less trustworthy (-18). 84 Source: Reuters 2025 Digital News Report Latest trends in media literacy Over half of the respondents remain concerned about their ability to tell what is true from what is false (58% globally, 46% in Western Europe). 85 Source: Reuters 2025 Digital News Report The European Democracy Shield 86 Safeguarding the integrity of the information space Strengthening democratic institutions, free and fair elections, and free media Boosting societal resilience and citizens’ engagement Launched on 12 November 2025: Strategic approach to safeguard, strengthen and promote democracy in the EU Series of actions, progressively rolled out by 2027 Empowering Europeans in whole-of-society approach Media literacy in the Democracy Shield Promoting media, digital literacy and critical thinking: strengthen awareness of information integrity risks empower citizens of all ages across all communities to navigate the information environment, access reliable information, exercise their democratic rights and engage in democratic processes. 87 Media literacy in the Democracy Shield Package of measures on media literacy to be rolled out as par of the Media Resilience Programme: new actions reaching all age categories linking media literacy to societal resilience against disinformation & to preparedness focusing on new trends in terms of media consumption supporting Member States that face particular challenges. 88 Media literacy in the Democracy Shield Reinforcing the European media literacy community: 89 Media literacy expert group National governments to cooperate more closely & expand mandate to preparedness Network of independent experts Get fresh insights from various disciplines to support innovative, evidencebased media literacy policy Other stakeholders Build on existing best practices & success stories e.g. European Digital Media Observatory (EDMO) 04.12.2025 From Efficiency to Democratic Effectiveness: Rethinking AI in Deliberative Processes 96 KT4D: Testing the Digital Democracy Lab and Demonstrator Platform Brussels Madrid Krakow Dublin Brussels Digital Democracy Lab: 1-day deliberative workshop with participants critically engaging with AI in democratic decision-making. Digital Democracy Demonstrator: a digital platform integrating LLMs to facilitate and test deliberative processes, while being as transparent as possible. Purpose: To explore, not optimize democracy through AI –by embedding meaningful frictions into AI interactions. Facilitation matters more than technology Civic understanding > technical skills Friction improves reflection and agency True transparency is the basis of trust AI is prompt for dialogue, not a decision-maker 04.12.2025 From Efficiency to Democratic Effectiveness: Rethinking AI in Deliberative Processes 97 What we observed in the Digital Democracy Lab Brussels Algorithmic accountability and transparency = an ongoing relationship, not a report Real-time contestability Human agency over the technologyy Value-governance: align technology with democratic purpose 04.12.2025 From Efficiency to Democratic Effectiveness: Rethinking AI in Deliberative Processes 98 Democracy-in-the-loop: a new development and governance approach Brussels What is the added value for democratic effectiveness? What does the AI replace? Can participants understand how it works? Can facilitators (or participants) challenge it in real time? Can friction prevent artificial consensus? 04.12.2025 From Efficiency to Democratic Effectiveness: Rethinking AI in Deliberative Processes 99 Before introducing AI in democratic processes, ask: Brussels Start with the democratic purpose, not technology Identify the substitution effect Build friction into democratic technologies We must design technologies that slow us down, open things up, and keep power in the hands of the people 04.12.2025 From Efficiency to Democratic Effectiveness: Rethinking AI in Deliberative Processes 100 Looking forward: three recommendations Brussels Thank you Elizabeth Calderón Lüning Director of Programmes at Democratic Society demsoc.org eliz[email protected] AI for Public Deliberation Prof. Anna De Liddo The Open University, Knowledge Media Institute, UK ORBIS Project https://orbis-project.eu/ 04-12-2025, Bruxelles Æ, BłGDATAD*Ğ•Ś' '*'• Athena KnowledgeMediaInstitute Anna De Liddo Professor of Human-Computer Interaction Lady Kitty Chisholm PhD Student Belief Change in Adults at Work Alberto Ardito Senior Web Developer Retno Lasarati PhD Student Explainable AI Lucas Anastasiou PhD Student Argument Mining Riccardo Pala Web Developer Michelle Bachler Senior Project Officer Problem 4 Profound luck of institutions, norms, methods and tools to engage the public in healthy deliberation on issues of public concern 112 TheORBISApproach:AHybridHuman-AI Collaboration ORBIS employs a "Human-on-the-Loop" (HOTL) approach. Al augments, rather than replaces, human decision-making, ensuring that data-driven insights are aligned with human reasoning, values, and ethical oversight. TheORBISToolkit Feedback Aggregator Argumentation Mining Policy Composition &Collective Decision Making Explanation Generator Structured Insights & Visualizations DELIBERATIVE PLATFORMS third party platforms BCause Democratic Reflection PolisORBIS ORBIS AI-boosted toolkit to augment deliberative democracy Bcause platform and Democratic Reflection Rob deMatt-PoliMi ReImagineEurope CNRS-NovelCore Università della Svizzera CEPS,PDE, DCF ✓equally accessible, ✓more structured, ✓and higher data quality than common Social Media Reasoning for change A Structured and Decentralised Discussion System for Distributed Decision Making. Explore Lorem ipsum dolor sit amet, consectetur adipiscing elit. Sed non sem cursus, commodo dui vel, interdum est. Aliquam malesuada rutrum neque, in laoreet lacus facilisis quis. Vai alla sezione Reflect Lorem ipsum dolor sit amet, consectetur adipiscing elit. Sed non sem cursus, commodo dui vel, interdum est. Aliquam malesuada rutrum neque, in laoreet lacus facilisis quis. Vai alla sezione BCAUSE REASONING FOR CHANGE Login Signin EXPLORE REFLECT DISCUSS DECIDE bcause.app A novel and cutting-edge technological solution for online deliberation, which is: BCAUSE REASONING FOR CHANGE yş How CanManmade Climate Change BeReversed? Is ^^ thateven possible? TIMELINE Briefly enter yourpositionhere I É ( ) 1 yearsago iMrs.SueKlein: li may be technically possibleto reverse climate changebut is gettingharder yearby year )1yearsago Give your opinion Refect (6) ' andother19 people Give your opinion tRetient (2) '. CONS 1 yearsago MrsSueKlein: C 0 If we identify the negative feedback loops and break them,then we have a chance of slowing or even reversing climate change Refiect (3) 1yeasago PROS Climatechangeis accelerating-dueto feedbackloops-muchfasterthat whathumanity can-or is willingto-decelerate it.Unfortunately isnotfeasible withcurrent technology andother919 people 06/05/21 EXPLORETHE ANALYTIŒ A, BIGDATAD”' D E M D C R A G Y T A SK H¢›w Can Manmade Climate ChangeBe Reversed7is that possi ie Iam sorry but NO Almost the wholehuman civilisation runsof fossil fuels pumped from thedeep carbon cycleto the fast carbon cycle. It will reverse naturally over several million years, not the short-term which we can benefit from right now CLOSE Isomewhat agree 11/05/2022 *S This is a very polarized position It is clear I can trust this AI,BIG DATAD ** s'”• DEMOCRACYTASN FORCE HowCanManmade Climate ChangeBe Reversed? Is that evenpossible? Iamsorry but NO. Almostthe whole humancivilisation runsof fossil fuels pumped from the deepcarbon cycle to thefast carboncycle. It will reverse naturally over several mlllion years,not the shon-term whlch we canbenefit from right now. To whatextenddo youagree that NEXT 2 consumers need to demandmore from the same organizations. Weneed to go to all levels of government anddemand that these same organizations give us alternatives in the form of electric andhvdro cars while also Today 17 https://bcause.app/ Structuring the Dialogue: From Raw Transcripts to Argument Maps bridge between synchronous events and asynchronous online debates by providing import and Human-AI analysis of transcript of live events BCAUSE REASONING FOR CHANGE The Curation Workflow: Putting Humans in Control 18 Al-generated argument structures are presented as initial proposals, not final determinations. Human moderators retain full control to refine, correct,and validate the system's output. Control Parameters The system allows to influence the number of generated posts, the ratio of generated arguments per position and the bias towards contra-arguments Approval &Historical Log Each argument node has an approval mechanism.The system retains a log of all curatorial actions to ensure transparency and accountability. 19 Bridgingthe gap between local issues and policy, ORBIS facilitates 'Geo-Deliberation Geo-Deliberation with Rob De Mat in Milan 20 Transforming static map data into a starting point for collective problemsolving. 13 bcause--conflict-detection-4z1ky6a5.web.app/discussions/-ORLhn1IJQpgnfz-KvOu?target=-ORLhn1IJQpgnfz-KvOu „@Wholesale Clothin... OpenUniversity :... Debate Hub-local av DemocraticREPLAY Member Online -... OLiteMap Bookmark... Cards of Clarity DOMINATING VOICE ONEORMORElMDlVDUALSWOMOPOLZE •We'veheard a lot of valuable input on these topics. Let's take a moment to invite other participants to share their thoughts, questions, or alternative viewpoints. ' ' “ “ “ “"“ “ “' “ '“ QUIET CORNERS Briefly enter your position here PGR-Manager-Op... OLiteMap Bookmark... Transcript data, imported by: Lucas Abased on1 evidences The European Bauhaus can provide support to smaller entities through communication, networking, and connections to resources and venues. Give o r op non R e f u t (0e) Idisagree 6months ago ago I agree transcript data basedon2 evidences 28/05/25 2 ArgumeMs mo The European Bauhaus offers communicatio n assistance and networking opportunities t o help smaller entities. Reflect 'e iden! 6months ago ago *d’’a based on2 Today » @ AllBookmarks ANALYTICS TODAY O Ensuring a Healthy Dialogue with Conflict Mediation Tools The goalis not to suppress disagreement, whichis vital for deliberation,but to ensureit remains manageablethrough reasonable argumentation.These tools help moderators foster an environment where all participants can express opinions freely yet respectfully. Human Moderator AI Moderator Can use the cards as a guide for manual intervention and facilitation. Can be configured to automatically flag patterns and suggest interventions, such as pop-ups with critical thinking questions. USI: Universita’ Della Svizzera Italiana After the deliberation, BCause provides an advanced reporting system that empowers moderators and policymakers to generate comprehensive, interactive dashboards for analyzing the discussion. Drag-and-drop widget reorganization Resizable widgets to emphasize specific analytics Individual widget export (.png/.pdf) Full dashboard export to Google Docs for further refinement Synthesizing Collective Intelligence: The Analytics Dashboard The system features a modular approach with customizable widgets that can be tailored to specific analytical needs, creating temporal snapshots that capture the state of deliberation at specific moments. Analytics in Focus: Participation and Content Statistics Analytics in Focus: Al-Powered Insights Analytics in Focus: Al-Powered Insights The Argument Network provides a force-directed graph of the entire discussion, moving beyond a linear view to visualize the deep logical structure. It allows stakeholders to identify clusters of agreement, key points of conflict, and the central arguments shaping the consensus. By making the complex interplay of ideas visible and understandable, the ORBIS toolkit empowers communities and policymakers to engage in more informed, evidence-based, and ultimately more effective democratic deliberation. Democratic Reflection transforms passive content consumption (like watching a debate) into active critical engagement. It moves beyond simple "like/dislike" interactions to promote deeper, more nuanced thinking. 1. Engage An interactivereflective layer is overlaid onto live or recorded video content. 2. Reflect As users watch, the platform presents digital reflection cards with carefully crafted statements, acting as cognitive nudges. 3.Analyze Thesystem captures engagement patterns and generatespersonalized analytics, visualizingthe user's own reflection journey. The platform prompts viewers to examine the content through specific analytical lenseseffectively guiding them from passive watching to active critical engagement. Deepening Engagement Through Democratic Reflection 35 Democratic Reflectiofi’s Ifiteractive Experiefice Speaker A I Lisbon Deliberative Focus Groups (NOVA) Transcrição O:OO 0:01 Penso que a abertura do novo aeroporto Luís de Camões em Alcochete é muito necessária. O espaço aéreo de Lisboa está sobrecarregado, dada o aumento constante do número de turistas e residentes que entram e saem da cidade. O atual aeroporto Humberto Delgado, por si só, já não consegue suportar todo o tráfego aéreo. Mas isso nãosignifica que o antigo aeroporto deva ser encerrado. Na minha opinião, Enquanto o novo aeroporto de Alcochete alivia o tráfego aéreo, o aeroporto Humberto Delgado deve ser mantido, ainda com capacidade reduzida. Em vez de um encerramento total, o Humberto Delgado pode funcionar como um aeroporto premium, de baixo tráfego, centrado em rotas internacionais de alto valor, voos governamentais, jatos empresariais e viagens diplomáticas ou culturais. Este modelo reduz a pressão sobre o nova aeroporto de Alcochete e, ao mesma tempo, mantinha o acesso direto ao centro da cidade para serviços críticos. Na verdade, este modelo já funciona bem em muitas outras cidades que adotaram esta solução especializada de dois fl:00 lz. Cronologia de Reflexões Durante adiscussão, os espetadores clicaram nos seguintes cartões de reflexão. Clique em cada barra para deslocar para aparte correspondenteda transcrição àesquerda. @Painel @Relatório @ Inquéritos •W Português @15sec 0.00 0150:30 0.451:001151:301:452:00 2152:302.453.00 3:153.30 3.45 4.00 4:154:30 4:45 5.00 5.15530 5456:00 6156.30 6457 007.157:307:45800 8:158:30 8.45 9:00 ■Iagree with this ■Idisagree, but Iunderstand his point ■Idisagree withthis ■Thispoint makes sense to me ■This point isn't well supported ■This is nonsensical ■This makes me angry ■This gets me thinking ■This makes me happy ■I think he is biased Temas de Discussão Estes são os temas identificados à medida que se desenrolaram durante a discussão ■Many people could agree with him ■Ithink he is reasonable OManutenção do Humberto D Impacto Económico e Emprc Odebate central ése oaeroporto Humberto Delgado deveser totalmente encerrado após a abertura de Alcochete ou se deve ser mantido com funções reduzidas, como um aeroporto premium. Innovation for a Safer Digital World Safety Tech Shaping Trusted Spaces in the Age of AI Zixuan Fu I am Zixuan, from a company called PUBLIC, where we work a lot with UK and European governments on complex digital and technology challenges. I lead our research and policy advisory work on AI and digital governance. We work closely with technologists, innovators, practitioners, and policymakers to understand evolving online threats and support effective technological and policy solutions, particularly in areas relating to trust, safety, and harm in online environments - including producing our annual International state of Safety Tech report. [email protected] Working with government, industry, researchers and civil society organisations to build safer and trusted online spaces Annual GovTech Summit with 2,000+ attendees Gov-funded innovation programme 25+ Online & AI safety projects since 2020 40 Organisations regularly engaged 600~900 International Safety Tech innovators mapped From GenAI to embodied intelligence, tech-enabled threats are increasingly challenging trust and security across sectors and society Public concerns about trust remain high 90.4% of people concerned about deepfakes 87.4% fear deepfakes could manipulate election results AI-generated child sexual abuse material has risen at an alarming rate More than doubling within a single year Shifting tactics towards digital threats A new cyber-fraud incident occurs Every 2 minute Data sources: [1]The Alan Turing Institute, Behind the Deepfake: 8% Create; 90% Concerned, 2024. [2] Internet Watch Foundation, AI imagery getting more ‘extreme’ as IWF welcomes new rules allowing thorough testing of AI tools, 2025. [3] Cifas, Fraudscape 2025. Risk factors driving GenAI-enabled online threats Risk factors GenAI-enabled / facilitated threats Scale Realism & Fidelity Normalisation Persistent memory Speed Reasoning Opacity Misaligned intent Autonomy behaviours Bias amplification ●Misinformation & Disinformation ●Synthetic Media & Deepfakes ●Social Engineering ●Impersonation & Identity Fraud ●Data Poisoning ●Model Manipulation ●Privacy Leaks ●Automated Cyber-attacks ●Reputation & Institutional Harm Source: PUBLIC Analysis System -wide solutions Tracing, locating and removing illegal content Platform moderation Monitoring, detecting, flagging harmful behaviours and content Age -related online safety Ensuring age -appropriate experience, age assurance and digital identify Device -level protection User, parental or device - based security products Network filtering Filtering, blocking and flagging harmful or illegal content at the network level Information integrity Detecting and disrupting false, misleading, or harmful narratives Physical -domain safety Identifying, preventing, and responding to harmful realworld incidents Professional services Advisory support with implementing technical solutions AI safety, security and assurance Our research develop a Safety Tech taxonomy: Source: PUBLIC Analysis “Safety Tech providers develop technologies or solutions to facilitate safer online experiences, and protect users from harmful content, contact or conduct.” —UK Government Department for Science, Innovation and Technology Safety Tech - Key Innovation Pillar Safety Tech in the Wider Ecosystem A collaborative ecosystem is essential for effective online safety and digital governance Policy & Regulation (Govts, Intl’n bodies) Civil Society & End Users Tech Industry (Platforms, AI developers) Research & Academia (Universities, Think Tanks) Safety Tech Innovation Graphic source: PUBLIC and Paladin Capital Group, International State of Safety Tech 2024 Challenges for Safety Tech 1. Balancing Privacy and Safety Navigating the complex line between protecting users and upholding fundamental rights, such as privacy and freedom of expression 2. Pace of Policy/Regulation vs. Innovation Developing agile policy frameworks that keep up with rapid technological advancements without stifling innovation 3. Technical barriers slowing down adoption Several technical obstacles exists, including limited access to high-quality data for training and testing, a nascent market, and early-stage technologies with gaps in evidence and independent evaluation Source: PUBLIC Analysis From Research to Action What does this mean … … for policymakers & regulators: Provide dedicated funding - such as via grant and challenge programmes - to support early-stage Safety Tech R&D, prototyping, and pilot deployments. Create safe spaces - such as regulatory sandboxes, innovation test beds - to trial new detection, moderation, and risk-assessment technologies under controlled conditions. … for industry: Position trust and safety as core product differentiators. Enable technical interoperability so solutions can integrate and reinforce each other. … for researchers: Horizon scanning to identify emerging risks and novel technological solutions, producing evidence-based insights that inform policy-making. Champion cross-sector collaborations Open Discussion + Q&A For those who join us online, please use the Q&A box to pose your questions 153