Research & writing.

Maps sensemaking tools along two dimensions: which stage of the interpretive process a tool addresses, and where its interpretive intelligence sits. Underneath is a distinction between inductive work, which extracts patterns at scale, and abductive work, which builds a good-enough explanation of incomplete and ambiguous data. AI is strong at the first; at the second it can generate candidate explanations but cannot reliably or accountably judge which one fits. Across five domains of practice the imbalance is systematic - signal detection and translation are densely served, frame construction and interpretive integration are not - and the paper argues that the arrival of these tools is itself changing what gets captured.

Examines a youth consultation run in Yemen under conditions that make conventional participation difficult: protracted conflict, territorial fragmentation, and severe constraints on convening people physically. The method paired a WhatsApp-based chatbot with AI-assisted qualitative sensemaking, and the report assesses what that did and did not resolve for inclusion, scale and security.

Argues that PeaceTech's recurring failures are structural rather than accidental. Tools built to empower communities can enable surveillance once crowdsourced data exposes vulnerable groups; early-warning alerts frequently produce visibility without response; and corporate control of cloud and satellite infrastructure puts a public good in private hands. Proposes guardrails - service continuity, community data ownership, algorithmic accountability, and response protocols tied to funding.

A study of 22 facilitators worldwide, setting out where AI could genuinely support group conversation rather than substitute for it. Identifies five underserved opportunities - scaling access, dynamic learning, live synthesis and sensemaking, futurecasting, and public sensemaking outputs - and separates transformative, generative and deliberative processes, since which tool helps depends on what a session is actually trying to do.

Proposes a context-sensitive taxonomy of digital deliberation tools for conflict-affected environments, turning on scale, digital literacy, inclusivity, security and how deeply AI is integrated. Drawing on UN and civil society cases, it makes the argument against one-size-fits-all platforms and for hybrid models that keep human facilitation in the loop where trust and legitimacy are at stake.

A qualitative study of digitally enhanced dialogue in Yemen and Libya. It distinguishes “sincere” affordances, which push participants to engage with the conflict as it is, from “subjunctive” ones, which open up how it could be - and argues that future-oriented peacemaking needs both. More data and evidence alone can harden a deadlock rather than break it.

The journal treatment of how CMI builds foresight into dialogue and mediation, using data analysis and visualisation to support broad participatory analysis. Argues that interactive visual aids help stakeholders make sense of a situation collectively and can loosen entrenched positions, so that shared future visions become possible even where a process has stalled.

The shorter practitioner version of the same argument, written a year earlier. Sets out the methodology CMI developed on the Inclus platform for mapping where stakeholders currently stand and then working with them toward alternative scenarios - a move from reactive problem-solving to proactive visioning.

Peacebuilding tends to under-address the emotional and psychological dimensions of conflict. This piece argues that behavioural science offers usable instruments for them - social norm nudging and narrative shifting among them - and makes the case for pairing behavioural expertise with practitioners when dialogue and reconciliation processes are designed.

An account of a dialogue that brought Ukrainians with opposing political views into contact across the line of conflict, in a hybrid online and offline format. Participants crowdsourced discussion topics anonymously in a closed group, voted on priorities, then met in facilitated video conversations - a design that surfaced identity-based issues conventional formats struggled with, while remaining skewed toward younger participants and exposed to disinformation.

Examines what digital tools can and cannot do about the dilemmas of inclusion in peace processes - the trade-offs that appear as soon as participation is widened beyond the small set of actors who reach the negotiation table.