Deliberative Technologies in Youth, Peace & Security

PAPER OF THE WEEK, 20 Jul 2026

Lena Slachmuijlder | Toda Peace Institute - TRANSCEND Media Service

Promise, Practice, and the Uptake Gap

Abstract

17 Jun 2026 – This paper draws on experimentation to ask a grounded question:

When peacebuilders and Youth, Peace, and Security (YPS) coalition leaders put AI-enabled deliberative tools into practice, what do they enable, what obstacles do they face, and how can we take forward their lessons learned?

This paper does three things.

First, it shares insights from practice—four concrete use cases showing how tools were selected, designed, framed, and how participation actually unfolded.

Next it synthesizes 18 months of experimentation—highlighting both the promise and the friction points, especially the uptake gap that practitioners hit even when the tools ‘work’.

Finally, it offers grounded recommendations—the enabling conditions that most strongly determine whether deliberation widens listening, strengthens inclusion, and supports real sensemaking.

****************

1. Why deliberative technologies for advancing YPS

The African Union’s Digital Transformation Strategy 2020-2030 reads, “African Union Member States and YPS stakeholders should explore ways to harness digital technologies and build inclusive infrastructures for consultation, collective analysis, and collaborative decision making.”

 That call—issued in the continental 10-year reflection on YPS in Africa—comes after several years of experimentation by youth coalitions, civil society, and digital peacebuilding organizations trying to make participation more inclusive, safer, and more practically useful.

Across Africa, the last decade of YPS efforts has clarified that participation cannot remain a one-off event. In fragile and polarized settings, participation is expected to do more than create visibility; it is expected to help prevent violence, strengthen legitimacy, and build social cohesion under pressure. That requires a listening capacity that is repeatable, inclusive, and able to surface both shared priorities and real fault lines—not only a series of consultations that produce documents.

This is one reason frameworks like the Common Impact Framework (CIF) and related ‘vital signs’ approaches have gained traction in continental and national YPS conversations. The CIF stems from Search for Common Ground’s Peace Impact Framework (PIF), which tracks five vital signs of a healthy, just and peaceful society. This framework offers a shared structure for learning and accountability: a way to connect participation in peacebuilding policy and practice to signals about trust, safety, agency, access to resources, and responsiveness, and to track those signals over time rather than treating participation as a snapshot.

But a second reality is equally important: the tools most commonly used for youth participation still tend to over–represent those already closest to power including urban networks, highly educated youth, those who speak dominant languages, and those who can afford to travel or have access to institutional spaces. The result is not only exclusion; it is a loss of signal. Capital-city workshops often produce polished narratives, but they can miss the textured, region-specific experiences that determine whether the public views a policy as legitimate, whether they trust a coalition, and whether prevention efforts will stick.

AI-enabled deliberative technologies have entered this space as potential complements to existing participation approaches. They promise to widen who can participate, increase psychological safety, and make patterns legible at scale. Yet their contribution is not automatic. The key question is whether these tools can be used by practitioners in ways that genuinely expand the listening perimeter and strengthen collective sensemaking—without reproducing elite capture or turning into yet another extractive consultation layer.

Search for Common Ground supported YPS networks in using deliberative technologies across Africa and at the global level. This paper uses experiments to ask a practical question: When peacebuilders and leaders of the Youth, Peace, and Security (YPS) group use AI tools, what do they help with, what problems do they face, and how can we improve based on what they learn? This paper has three goals. First, it shares real-life examples by describing four cases that show how tools were selected, designed, and used, and how people participated. Next, it summarizes 18 months of testing, highlighting both the potential and the problems, especially the gap in using the tools even when they work well. Finally, it offers practical advice on the conditions that most effectively made discussions more inclusive and meaningful.

2. Value add of deliberative technologies

Deliberative technologies such as Pol.is, Remesh, and Talk to the City differ from conventional consultation tools in how they treat voice and disagreement. Instead of asking participants to respond to fixed survey options, these platforms can allow participants to:

• generate statements in their own words,
• respond anonymously or semi-anonymously,
• react to others’ ideas through structured interaction, and
• produce analytic clustering that shows where views converge and diverge.

In YPS contexts, several features matter in particular:

• Psychological safety. Anonymity (or reduced identity visibility) can lower fear of reprisal and social pressure, allowing more candid participation in politically sensitive environments. People respond to ideas rather than to the perceived status or identity of the person speaking.
• Collective reasoning at scale. Clustering helps reveal patterns such as shared priorities and fault lines without forcing majority votes or erasing minority perspectives.
• Inclusion through distribution channel fit. Tools that work through familiar social media or messaging channels (notably WhatsApp) and support voice input can reduce barriers related to literacy, language, and confidence. This matters when the goal is to reach youth beyond established networks.
• Reduced in–person facilitation bias. Because patterns emerge from aggregated responses and voting (rather than from who speaks most in a workshop), deliberative tools can reduce the extent to which outputs are shaped by in-person dominance dynamics.

At the same time, deliberative tools are not a substitute for political judgment or coalition leadership. They do not remove power asymmetries. They do not ‘solve’ legitimacy questions on their own. Their value depends on design choices (prompts, onboarding, channel selection) and on what conveners do after participation: interpretation, sensemaking, and responsible communication of results.

3. Use cases: What was tried, why, and what emerged

3.1    POL.IS – GLOBAL YPS CONSULTATION ACROSS 42 COUNTRIES

Purpose and convening group

A Search for Common Ground YPS Fellow from Afghanistan sought to gather perspectives from across youth-led networks in conflict-affected countries and diaspora communities from around the world. The focus was to gather these perspectives and turn them into a set of recommendations for the United States government about youth perspectives on effective American policy to engage young people to prevent violent extremism. Rather than sending out a pre-framed survey to which youth could respond, the Fellow chose to use Pol.is to trigger young people’s perspectives in their own words, and enable deliberation and interaction. By choosing Pol.is, the process would be inclusive, still legitimate, but with less constricted boundaries than a traditional survey.

Why Pol.is

The team selected Pol.is for its simplicity in use and ability to remain open for asynchronous participation, which is an asset when seeking to listen across time zones. It also produces a reporting format that translates relatively well into advocacy: clear maps of consensus and divergence and clustering that help distinguish broad agreement from contested claims. The platform’s user interface enabled the use of third-party translation tools, which also enabled greater multi-lingual participation in a unique shared platform.

Prompt and framing choices

The conveners tested alternative framings and selected a prompt aimed at eliciting implementable recommendations. The design challenge was to avoid a survey mindset: the question needed to be focused enough to generate policy-relevant insights, while still allowing participants to generate statements in their own terms rather than reacting to predefined options. Key to the design and use of Pol.is was the choice of the initial ‘seed statements’. These were chosen to hit several criteria: offering an accessible natural language, not-overly-generic policy language; presenting a diversity of perspectives thus giving permission for people to disagree with seed statements; articulating specific ideas, not multiple or general terms, to model the type of comments participants would themselves offer through their participation.

Participant outreach and onboarding

The link was circulated to 113 young people through dozens of youth networks, largely via private WhatsApp groups and direct messaging. This distribution choice used trusted channels, reduced the risk of trolling, and supported psychological safety. Participants were given a short orientation explaining how statement generation and voting would work and what anonymity meant in practice. The process ran within a defined participation time frame, with reminders to encourage voting participation (since in many Pol.is engagements, a small group generates many statements while the broader group contributes primarily through voting).

What emerged

Participants generated 61 statements and cast 857 votes. Strong convergence formed around several policy-relevant claims: the need for donors and policymakers to meet youth needs in conflict-affected contexts; the importance of approaches sensitive to youth grievances; and a call for strategies that treat youth as partners rather than as troublemakers. The deliberation also surfaced clear fault lines. A contested statement about the comparative effectiveness of military action versus youth-led peacebuilding divided participants sharply by opinion group. Another contentious theme concerned whether youth are actually involved in resolving conflicts in meaningful ways.

Practice insight

Conveners noted that anonymity and asynchronicity widened participation beyond established networks and enabled cross-border pattern recognition without expensive convening. But they also observed an early version of the uptake gap: translating outputs into policy narratives required careful interpretation, particularly when divergence appeared on politically charged issues.

3.2  TALK TO THE CITY – REACHING BEYOND ESTABLISHED YOUTH NETWORKS IN NIGERIA

Purpose and convening group

In Nigeria, the convening group included members of YPS coalition structures and partner organizations with trusted connections beyond the capital. They served as local connectors who could distribute participation invitations through credible channels. They chose to use the platform Talk to the City during the review of the first year of Nigeria’s YPS National Action Plan (NAP) to address a participation gap: Abuja-centric consultations risked under-representing youth from conflict-affected regions in the Middle Belt and Northeast

Why Talk to the City

The design problem here was not only scale but accessibility. They selected Talk to the City because of how it deploys a bot within WhatsApp which supports voice-based participation, reducing barriers linked to literacy, language, and the confidence required to write long responses. The choice also reflected the reality that in many contexts youth are more comfortable sending a short voice note in a familiar channel like WhatsApp rather than participating in a formal workshop or filling in a written survey.

Prompt and framing choices

The team designed prompts in plain language and sequenced from lived experience to priorities. The aim was to gather grounded insights that could inform the NAP while keeping participation low-friction. Conveners avoided technocratic questions and instead asked questions that could be answered naturally through voice notes, then later mapped to policy priorities through synthesis.

Participant outreach and onboarding

Distribution relied on coalition and partner channels: WhatsApp groups, local youth leaders, and community networks. The outreach model was intentionally trust-based rather than public broadcast. Participants received a short message explaining how to contribute (voice or text), what anonymity meant (no attribution in reporting), and how the data would be synthesized.

What emerged

Inputs surfaced concerns less visible in capital-focused consultations, including policing abuses, educational exclusion, and ethnic tension, often described with region-specific nuance. AI-assisted transcription/translation and clustering helped make patterns legible across diverse contributions. Organizers also observed a representational effect: constituency voice sometimes diverged from the assumptions of more formal youth networks, prompting internal coalition reflection.

Practice insight

In this case, the tool’s inclusion value was inseparable from the accessibility of the tool through WhatsApp. The design lesson was clear: widening listening beyond established networks often depends less on the sophistication of the platform than on the fit between channel, trust, and participant comfort.

3.3 REMESH – DRC YPS MID-TERM ASSESSMENT

 In the Democratic Republic of the Congo, researchers working with YPS coalition actors and the Ministry of Youth undertook a large-scale listening exercise to understand how young people perceived change over the past decade and what should be prioritized next.

Design choices

The team used a mixed set of tools: structured data collection via surveys and broad outreach channels, complemented by in-person focus groups to reach lower-connectivity settings. Remesh was added to provide a deliberative layer where participants could respond in their own words and react to others’ statements, in contrast to the largely multiple-choice answers to the pre-framed survey choices. Anonymity helped participants raise concerns more candidly in a politically sensitive environment.

What emerged

Remesh contributed a set of more personalized answers reflecting the experiences of young people in feeling that they were subject to political manipulation or invited into symbolic, yet ineffective, participation and consultation. Conveners valued the platform’s AI-enabled analysis ability to generate thematic synthesis while retaining ‘human flavour’. But conveners also noted that as the inputs became richer and more candid, the more careful they had to be about what they elevated and how they communicated findings, in order to preserve the safety of the anonymous participation. This related to ensuring that having ‘real quotes’ from participants did not override the more general trends.

3.4  REMESH – AU–EU YOUTH ADVOCACY RISK MAPPING (PPIRA)

 As part of a project aimed at strengthening advocacy collaboration across African and European Youth networks, the Search for Common Ground project teams conducted an online Participatory Protection and Inclusion Risk Analysis (PPIRA) with 51 youth across Africa and Europe, focusing on how participation can become risky or extractive and what protections enable safer advocacy.

Design choices

Remesh enabled an online PPIRA by combining structured questioning with anonymity and peer visibility. Separate English and French sessions reduced onboarding friction and supported facilitation. The question flow focused on how participants define meaningful participation, where tokenism and adultism constrain influence, what risks arise in advocacy contexts, and what practical mechanisms would make participation safer and more consequential.

What emerged

Participants articulated a consistent diagnosis of ‘participation without power’, distinguishing visibility from meaningful influence. They emphasized that safety and inclusion are not add-ons but design requirements for participation. Organizers benefited from real-time synthesis and post-session analysis; participants valued the ability to register agreement and divergence without exposure.

4. Lessons from 18 months of experimentation

Across the cases, deliberative tools repeatedly delivered three forms of value when used well:

1)   Widened listening beyond established networks—especially when channel fit was prioritized.

The Nigeria case demonstrated that accessibility and trust can be designed into participation through familiar channels and voice-based input. When this design logic held, participation widened beyond the most institutionally fluent youth.

2)    More candid participation in sensitive contexts.

Anonymity or reduced identity visibility consistently increased what participants were willing to surface, including concerns that can be risky to share in face-to-face settings.

3)    Faster, clearer sensemaking.

Clustering and synthesis helped conveners see patterns quickly such as seeing where there was convergence and divergence.The deliberative process supported coalition reflection and, in some cases, strengthened how the coalition framed its priorities. The AI-enabled summaries offered by Remesh also enabled the researchers to easily make sense of a large volume of qualitative input.

A second lesson is that the hard part often begins after participation. The tools can widen listening and generate clearer patterns quickly, but practitioners still have to make careful choices about interpretation, synthesis, and what gets elevated—especially when input is diverse, politically sensitive, or internally contested. The value is real, but it is not automatic: the impact depends on the craft of prompting, onboarding, facilitation, and share-back.

5. The uptake gap

Across the pilots, ‘uptake’ didn’t fail because people rejected the idea of deliberation. Many coalition leaders and civil society practitioners were excited by how quickly these tools could widen listening and surface patterns. The uptake gap showed up in specific, repeatable frictions which were mostly practical and political rather than technical.

5.1     THE CRAFT GAP: PROMPT DESIGN IS A SKILL, NOT A TEMPLATE

Teams learned quickly that a deliberative prompt is not a survey question. A prompt that is too broad produces noise; too narrow produces performative agreement. In the Pol.is consultation, the convening group tested seed comments to balance actionability for a policy audience with openness to youth-generated statements.

In Remesh sessions, question flow and length mattered: long sequences risked fatigue; overly abstract prompts triggered generic language with reduced usefulness.

5.2      THE ONBOARDING AND ACCESSIBILITY GAP: PARTICIPATION SCALES WITH FAMILIARITY

The Nigeria case highlights an inclusion principle that repeated across the experimentation: participation widened when the distribution channel (e.g. WhatsApp) matched what people already used and trusted. Because WhatsApp voice notes were convenient, they sparked greater inclusion. When onboarding was thin, or tools were unfamiliar, participation narrowed back toward the already-connected—English-speaking, tech-confident, urban youth. This is one of the clearest ways ‘efficient’ tools can unintentionally reproduce elite capture.

5.3      THE COALITION POLITICS GAP: PLURAL INPUT CAN DISRUPT INTERNAL CONSENSUS

When coalitions used these tools effectively, they often discovered that what leaders believed ‘the network agrees on’ did not match what youth outside the room were saying. That is valuable, but it can also create tension about representation and voice. It prompts the question of who speaks for the constituency, how they handle dissent, and whether the coalition has the relational maturity to treat divergence as signal rather than threat. In the above-mentioned Nigeria case, the in-person stakeholders welcomed the diverse perspectives gathered through a wider digital inclusion beyond the room. This may not always be the case, and merits attention to prepare people for an incoming diversity of perspectives.

5.4      THE INTERPRETATION AND SHARE-BACK GAP: SYNTHESIS STILL TAKES HUMAN TIME

Even when tools provide clustering and summaries, conveners still need time to interpret and package outputs responsibly. Remesh outputs, unlike the standard Pol.is report, may require more preparation precisely because of the advanced filtering and disaggregated analysis the tool provides. When coalitions lack time or budget for analysis and share-back, participants can shift from feeling empowered to feeling used.

Deliberation does not reduce the importance of feedback loops where participants get to see the output of the process. Rather, it can raise their expectations.

5.5      THE ACCESS AND LANGUAGE GAP: THE HARDEST CONTEXTS ARE STILL THE HARDEST

When practitioners attempted to introduce tools in settings affected by active conflict or heavy restriction, access constraints quickly became deciding factors. For example, efforts to use Pol.is with youth leaders in South Sudan had initially been included amongst the projects this paper intended to highlight, but lack of internet access made the effort unsuccessful, leading the team to resort to an off-line, sms-friendly compatible survey tool. In testing Remesh in Syria, participants need to be on a VPN in order to access the platform in many parts of the country, and in several instances the VPN appeared to alter some of the translation accuracy. These considerations shape whether these tools remain concentrated in relatively accessible settings rather than reaching those most excluded from conventional participation.

5.6      THE LEGITIMACY GAP: “IS THIS VALID?” KEEPS RETURNING

Some practitioners hesitated to cite outputs from deliberative tech-facilitated processes in formal spaces because outputs did not include demographic disaggregation or conventional markers of representativeness. Both the newness of the deliberative tool, juxtaposed with established norms around survey methodologies and disaggregation, created some doubts and hesitation about the validity of the outputs. It raised questions around what decision-makers consider credible, and what explanation do practitioners need to provide about recruitment, method, and limitations to prevent dismissal.

6. Recommendations

The recommendations below are intentionally limited to the conditions that repeatedly shaped success across the cases. They aim to highlight practical questions around whether deliberative tools widened listening, strengthened inclusion, and supported usable sensemaking.

6.1    TREAT DELIBERATION PROCESS DESIGN AS A CORE PART OF THE WORK, NOT A TECHNICAL ADD-ON

What you ask, how you sequence it, and how you seed participation are core to the methodology of using deliberative tools. Coalitions and conveners who gave time and iterated versions generated more relevant results. For example, they tested alternative framings, kept language plain, and matched the flow to the intended use of the outputs. Seed comments set the tone, give permission for open and frank diverse viewpoints, and reflect just one core idea. When the seed comments were done in a rush, they tended to be generic, vague, too wide, and not sufficiently diverse.

Implication: Build time for design and testing into YPS processes. If deliberation is inserted at the last minute, it will predictably underperform.

6.2      DESIGN FOR INCLUSION THROUGH DISTRIBUTION CHANNEL FIT AND TRUSTED MESSENGERS

The Nigeria example makes a broader point: inclusion is achieved less by ‘going digital’ and more by choosing familiar messaging channels, through invitations from trust coalition connectors, and enabling voice input. Had the in-room participants decided to simply publish a public link without involving trusted messengers, it is unlikely that the digital participation would have been robust.

Implication: If widening listening is the goal, treat channel fit and trusted outreach pathways as the primary inclusion lever—not an afterthought.

6.3      BUILD SHARED SENSEMAKING CAPACITY INSIDE COALITIONS, NOT ONLY ANALYSIS OUTPUTS

The tools can surface patterns quickly, but they do not resolve what those patterns mean. The outputs may raise new questions about how coalitions should respond when participants voice diverging views that conflict with internal assumptions. The experiments underline that the value of deliberation depends on coalitions being able to interpret results together and negotiate what to elevate without suppressing disagreement.

Implication: Pair deliberation with structured coalition sensemaking where leaders and connectors interpret outputs together and decide what they can responsibly carry forward. This can be based on a time or participation-level benchmark being reached, which then prompts reflection on the process next steps.

6.4      PLAN FOR PLURALITY IN ADVANCE: DON’T FORCE A SINGLE NARRATIVE WHERE IT DOESN’T EXIST

An underlying, not always visible friction was the readiness of the conveners of these deliberative processes to accept, process and then advance the pluralism and divergence that emerged. Deliberative or listening tools will not ‘deliver’ consensus and credibility; rather, they surface plurality in a way that may require further stages of dialogue, prioritization and constituency building.

Implication: Agree upfront how you will represent disagreement in reporting and advocacy, including what counts as a priority and how minority perspectives will be held.

6.5  MAKE FEEDBACK LOOPS NON-NEGOTIABLE

A recurring risk is that deliberation is pitched as ‘easier research’, being viewed as simply faster extraction. Yet the same research fatigue exists. When participants do not see how conveners synthesized and used their inputs, disillusionment grows—especially when participation asked for candid or sensitive contributions.

Implication: Commit to share-back as part of the method. If resources are limited, choose a lighter participation cycle that you can close responsibly rather than a larger cycle that you cannot.

6.6  ESTABLISH CREDIBILITY PRACTICES THAT MATCH THE REALITY OF YPS SETTINGS

Deliberative outputs are often challenged on legitimacy grounds—especially when demographic detail is limited. The answer is not to overclaim rigour, but to become disciplined about transparency: who was reached, through what channels, what the tool does, what anonymity means, and what limitations apply.

Implication: Normalize short ‘method notes’ for deliberative exercises so results can be shared in a way that clarifies the methodology and its limitations, so that their value is clearly understood.

6.7  TREAT ACCESS, LANGUAGE, AND SAFETY CONSTRAINTS AS DESIGN CONSTRAINTS, NOT OBSTACLES

The experimentation showed that the contexts where inclusion is most urgent are often those where platform access and language support are weakest. Similarly, anonymity helps, but safety requires additional safeguards: clarity about consent, careful handling of sensitive themes, and decisions about what is safe to circulate.

Implication: Select tools and participation formats based on the access and risk environment. Where constraints are high, hybrid approaches may be more realistic than fully online deliberation.

Conclusion

AI-enabled deliberative technologies are not a silver bullet for youth participation in YPS. But the experimentation suggests they can add real value when used with intention: widening who is heard, increasing candour where safety is a concern, and making patterns legible fast enough to support coalition sensemaking.

The deeper lesson is that outcomes depend less on the tool and more on the enabling conditions around it: design craft, channel fit, trusted outreach, shared interpretation, honest handling of plurality, and responsible feedback loops. The continental call to ‘harness digital technologies’ is therefore best read as an invitation to build inclusive participation infrastructure—not simply to run more consultations. If Africa’s next phase of YPS is to be more participatory and more accountable, the question is not whether deliberative tools exist, but whether the ecosystem invests in the practices that make them trustworthy, inclusive, and usable in the work of peacebuilding.

DOWNLOAD PDF FILE:

Deliberative-Technologies-in-Youth-Peace-Security

_______________________________________________

Lena Slachmuijlder works at the intersection of technology, peace, and conflict. She spent a decade with independent media in Africa before leading two decades of peacebuilding initiatives across Africa, Asia, and the Middle East with Search for Common Ground, where she headed the organization’s global program portfolio and championed the field of digital peacebuilding. Since 2023, she co-chairs the Council on Tech and Social Cohesion, a global network of technologists, peacebuilders, researchers, and policy influencers dedicated to steering technology toward social cohesion rather than polarization and violence. She supports the piloting of AI-enabled deliberative technologies to support collective decision-making, particularly in the Youth, Peace and Security space. A Senior Practitioner Fellow at the University of Southern California’s Neely Center for Ethical Leadership and Decision Making, she is a graduate of Stanford University and Amsterdam’s THNK School of Creative Leadership, and is based in Brussels.

The author would like to thank Lavina Oluoch, Omar Salem, Rachel Walsh Taza, Saji Prelis, Solvi Karlsson, Eoin O’Leary, and Yahya Qanie of Search for Common Ground, and Marie Rose Tshite, for generously sharing their experiences, insights, and reflections throughout the experimentation and use cases documented in this report.

The Toda Peace Institute is an independent, nonpartisan institute committed to advancing a more just and peaceful world through policy-oriented peace research and practice. The Institute commissions evidence-based research, convenes multi-track and multi-disciplinary problem-solving workshops and seminars, and promotes dialogue across ethnic, cultural, religious and political divides. It catalyses practical, policy-oriented conversations between theoretical experts, practitioners, policymakers and civil society leaders in order to discern innovative and creative solutions to the major problems confronting the world in the twenty-first century (see www.toda.org for more information).

PAPER OF THE WEEK STAYS POSTED FOR 2 WEEKS BEFORE BEING ARCHIVED

This article originally appeared on Transcend Media Service (TMS) on 20 Jul 2026.

Anticopyright: Editorials and articles originated on TMS may be freely reprinted, disseminated, translated and used as background material, provided an acknowledgement and link to the source, TMS: Deliberative Technologies in Youth, Peace & Security, is included. Thank you.

If you enjoyed this article, please donate to TMS to join the growing list of TMS Supporters.

Share this article: email mastodon facebook 🔗 copy link

This work is licensed under a CC BY-NC 4.0 License.

There are no comments so far.