Geneva Call Deed of Commitment

The Geneva-based mechanism, launched in 2000, by which armed groups and de facto or provisional authorities publicly commit to specific humanitarian norms through standardized deeds witnessed by Geneva Call and held by the Republic and Canton of Geneva.

Context

Geneva Call emerged from a practical legal gap. States can sign treaties. Armed groups cannot. Yet in many non-international armed conflicts, the armed group controls territory, commands fighters, runs detention sites, plants mines, recruits children, taxes communities, blocks health care, or governs food movement. The people exposed to those choices cannot wait for a treaty architecture that only states can enter.

The Deed of Commitment mechanism was Geneva Call’s answer to that gap. It gives an armed group, or a de facto or provisional authority, a way to accept a defined humanitarian norm publicly without becoming a treaty party. Geneva Call witnesses the deed, supports implementation, and monitors conduct. The Republic and Canton of Geneva receives a copy and acts as custodian. The legal design is narrow: the deed is a unilateral declaration by the signatory, not an agreement between political equals, and Geneva Call says the signature does not change the actor’s legal status.

The case matters because it is not only a document form. It is a running institutional practice: long-term contact before signature, fixed texts, political and military leadership signatures, implementation planning, training, reporting, allegation handling, and a public archive. Practitioners cite it when they need an example of humanitarian engagement with armed actors that is more than conversation and less than recognition.

What Was Tried

Geneva Call built the mechanism one theme at a time.

The first deed, launched in 2000, covered adherence to a total ban on anti-personnel mines and cooperation in mine action. The choice was not accidental. Anti-personnel mine use is observable, the 1997 Ottawa Convention had created a strong state treaty norm, and many of the remaining users were armed non-state actors who had no route into that treaty system.

The second deed, launched in 2010, covered protection of children from the effects of armed conflict. The third, launched in 2012, covered prohibition of sexual violence in armed conflict and elimination of gender discrimination. The fourth, launched in 2018, covered protection of health care in armed conflict. The fifth, launched in 2021, covered prevention of starvation and conflict-related food insecurity. The sequence shows the mechanism’s logic: start with a specific norm, consult legal and humanitarian experts, then ask an armed actor to sign a text it cannot rewrite.

The signature architecture is deliberately formal. Geneva Call’s public explanation says the deed is signed by the actor’s political and military leadership, countersigned by Geneva Call as witness and implementation partner, and also signed by the Government of the Republic and Canton of Geneva as custodian. The ceremony traditionally takes place in Geneva’s City Hall, in the Alabama Room, where the first Geneva Convention was adopted in 1864. That setting gives the act weight without making it a treaty.

The pre-signature criteria are as important as the text. Geneva Call names two essential conditions: a clear command chain capable of carrying orders through the organization, and direct, unimpeded access for Geneva Call staff to territory controlled by the signatory. If those conditions are not met, engagement may continue, but Geneva Call says it should not produce a deed. This is the case’s strongest operational discipline: a signature by a leadership that cannot command fighters is not implementation.

After signature, the deed is tied to an implementation plan. The plan names measures the signatory must take, the support Geneva Call will provide, and time frames for reporting and revision. Alleged violations trigger verification through several channels and confidential bilateral dialogue. Confirmed violations are first handled through remedial requests to the signatory’s leadership; Geneva Call reserves public reporting, suspension of dialogue, and possible repudiation for the most serious cases after other routes have failed.

What Worked

The mechanism created a route for armed actors to own norms they did not help draft. That is the central achievement. Geneva Call’s public archive, consulted in May 2026, lists 54 anti-personnel mine deeds, 31 child-protection and education deeds, 26 sexual-violence and gender-discrimination deeds, six health-care deeds, and one starvation and food-insecurity deed. Some actors sign more than one deed, so the totals should not be read as distinct organizations, but the spread across countries and themes is still unusual in armed-actor engagement.

The anti-personnel mine deed supplied the early proof of concept. Pascal Bongard’s 2008 review describes the landmine work as an attempt to include armed groups that could not accede to the Ottawa Convention but whose behavior determined whether civilians remained exposed to mines. The later Bongard and Somer analysis in the International Review of the Red Cross treats the mine deed as evidence that alternative monitoring mechanisms can improve compliance with at least some humanitarian norms.

The fixed-text design also worked. Because the deed cannot be modified, the signatory does not bargain down the norm clause by clause. The negotiation happens around ownership, internal adoption, implementation measures, and monitoring access. That matters for norms where partial acceptance would be destructive. A child-protection deed that weakens age rules, or a sexual-violence deed that leaves disciplinary procedures vague, would damage the point of signature.

The third-party custody structure gave the mechanism a recognizable non-endorsement frame. The Alabama Room ceremony, Geneva Call’s witness role, and the Canton’s custodian role make the act public and solemn. At the same time, the deed text and Geneva Call’s explanation insist that signature does not confer legal or political recognition. That combination is rare. Many humanitarian contacts with armed actors are either too quiet to defend or too public to keep from becoming status transactions.

The mechanism also produced learning across themes. The child-protection work forced Geneva Call to confront the difference between a mine ban and a more socially embedded problem: children may be recruited, used, associated with armed actors, or protected by them under very different conditions. The sexual-violence and gender-discrimination deed pushed the model into command responsibility, internal policy, women’s participation, training, and local civil-society monitoring. The health-care and starvation deeds extended the mechanism into areas where humanitarian access, service provision, and conduct of hostilities sit close together.

What Did Not

The deed mechanism did not erase the political problem of engaging armed actors. States can still read a signature ceremony as legitimation. Armed actors can still present the ceremony internally as recognition. Donors can still worry that their funds underwrite contact with designated or sanctioned actors. The legal-status clause narrows that problem; it doesn’t make the problem disappear.

The public archive can also overstate compliance if read carelessly. A signed deed proves a public commitment, not uniform conduct by every commander and unit. Geneva Call’s own monitoring procedure recognizes this. It treats alleged violations as expected facts of practice, not as surprises that invalidate the mechanism. The hard question is whether the actor investigates, remedies, trains, disciplines, and keeps the channel open after bad information arrives.

Fragmentation is a recurring stress point. Hichem Khadhraoui’s 2019 account of Geneva Call’s work describes groups that split, mutate, join umbrella formations, or lose stable leadership during protracted conflict. A deed signed by one formation may not travel cleanly into a splinter group. An allied formation may not share the same command discipline. A political office may remain committed while field commanders drift. The mechanism can’t follow those mutations unless contact is broader than a single senior signature.

Thematic uptake is uneven. Anti-personnel mines, child protection, and sexual violence account for most listed deeds. Health care and starvation have far fewer signatories in the public archive. That may reflect newer deed texts, harder monitoring questions, more direct military relevance, or the difficulty of asking armed actors to make commitments that constrain territorial governance and survival practices. The unevenness is itself a lesson: not every norm that matters can be turned into a deed at the same rate.

The mechanism also depends on access. Geneva Call’s own criteria require direct and unimpeded access to the signatory’s controlled areas. In practice, states may deny that access, front lines may shift, security may deteriorate, and the group itself may restrict contact. When access fails, monitoring becomes indirect, and the deed risks becoming a public symbol with a weak observation tail.

What Practitioners Draw From It

Practitioners usually draw five lessons from the case.

First, a public humanitarian commitment by an armed actor needs an implementation spine. Signature is the beginning of the case, not the end. Training, command orders, internal codes of conduct, focal points, civil-society contact, field visits, and violation handling are what make the deed more than a photograph.

Second, the signatory must be able to carry the norm. Geneva Call’s command-chain and access criteria are useful beyond Geneva Call itself. They force the analyst to ask whether the person signing can actually cause fighters to stop laying mines, stop recruiting children, protect clinics, or change food-blockade behavior. If the answer is no, the signature is premature.

Third, the norm has to be concrete enough to monitor. The strongest deeds concern conduct that can be observed, reported, trained, and investigated. Broad promises to respect all humanitarian law may sound stronger but often give monitors less to work with.

Fourth, formal ceremony and non-recognition can coexist if the design is strict. Geneva Call uses a historically loaded room, signatures, custody, and public records, but holds the act within a unilateral-declaration frame. The case is one of the field’s better examples of public ritual being used to raise compliance stakes without turning the signatory into a political peer.

Fifth, monitoring has to reach below leadership. The fragmentation literature points in the same direction as field intuition: senior leaders matter, but mid-level commanders, local civil society, community elders, religious figures, women’s organizations, and local health actors may be the people who detect whether the commitment is becoming behavior.

Disputed Lessons

The first dispute is whether the mechanism causes compliance or records a willingness that already existed. Supporters point to training, implementation plans, peer pressure, and public accountability as behavior-changing features. Skeptics note that an actor willing to sign may already have strategic reasons to comply, while a more abusive actor may refuse the deed or sign for reputational reasons. The sound reading is narrower: the deed is a compliance instrument for actors whose incentives and command system are already within reach, not a device that turns an unwilling actor into a disciplined one.

The second dispute is transferability. Geneva Call’s model depends on its Swiss identity, long-term field relationships, legal expertise, donor confidence, the Canton of Geneva’s custodian role, and a particular non-endorsement posture. Other organizations can learn from the anatomy, but they can’t copy the case by printing a new pledge form. Without custody, monitoring access, and trust in the witness, a deed-like instrument becomes a signed promise with no institutional weight behind it.

The third dispute is recognition. Some critics treat any public ceremony with an armed actor as legitimating. Some practitioners treat recognition risk as manageable if the text, venue, witness role, and communications are disciplined. The case supports the second position only under strict conditions. When the signatory can plausibly implement the norm, the text is narrow, monitoring is real, and the public record states the legal-status boundary, the humanitarian gain can justify the ceremony. When those conditions fail, the same form becomes a stage for premature recognition.

Related Patterns

Sources

• Geneva Call’s Deeds of Commitment archive supplies the primary record of the five deed texts, signatory counts by theme, signature architecture, command-chain and access criteria, implementation planning, monitoring, and violation-handling process.
• Geneva Call’s “How we work” page gives the organization’s current explanation of humanitarian engagement with armed groups and de facto or provisional authorities, including its statement that the work does not change their legal status.
• Pascal Bongard, “Engaging Armed Non-state Actors in a Landmine Ban: A Review of Geneva Call’s Action, 2000–2007”, Journal of Mine Action, 2008. Bongard explains why the mine-ban deed was created and how it addressed actors outside the Ottawa Convention’s accession route.
• Pascal Bongard and Jonathan Somer, “Monitoring armed non-state actor compliance with humanitarian norms: a look at international mechanisms and the Geneva Call Deed of Commitment”, International Review of the Red Cross, 2011. The article is the core source for the monitoring-and-verification argument.
• Pascal Bongard and Ezequiel Heffes, “Engaging armed non-State actors on the prohibition of recruiting and using children in hostilities: Some reflections from Geneva Call’s experience”, International Review of the Red Cross, 2019. The article shows how the deed model changed when the theme shifted from mines to children and armed conflict.
• Aurélie Lamazière, “Engaging armed non-state actors on the prohibition of sexual violence in armed conflict”, Humanitarian Exchange, 2014. Lamazière documents the early sexual-violence and gender-discrimination deed, including consultation, training, command responsibility, and local monitoring concerns.
• Hichem Khadhraoui, “Fragmentation of armed non-State actors in protracted armed conflicts: Some practical experiences on how to ensure compliance with humanitarian norms”, International Review of the Red Cross, 2019. Khadhraoui’s account supplies the case’s strongest caution about splintering, coalition formation, and leadership turnover.
• Annyssa Bellal, Pascal Bongard, and Ezequiel Heffes, “From Words to Deeds: A Study of Armed Non-State Actors’ Practice and Interpretation of International Humanitarian and Human Rights Norms”, Geneva Academy of International Humanitarian Law and Human Rights, 2022. The research brief places the deed mechanism inside the wider evidence base on armed-actor interpretation of humanitarian norms.