Recent events and experiences, including last year’s developments in Afghanistan and the war in Ukraine, have prompted HOT—and the wider humanitarian mapping community—to revisit what can be ethically and practically done to help people caught in conflict within the context of OSM.
This is not a new question; many of us have participated in mapping areas affected by conflict to some degree or another, but to date HOT has not created a formal set of policies and guidelines specifically addressing risk and harm to people from mapping in situations of conflict.
In addition to generating positive outcomes, data can also contribute to unintended risk and harm to people. In peaceful circumstances, we must still be careful not to harm others by, for example, revealing private and/or personally identifiable information (PII). The risks that arise from the kind of data that normally results from remote mapping campaigns—such as street and building traces—are well-understood, and there is a fairly robust consensus in the aid sector that the potential for harm is limited. Therefore organizations, OSM communities, and other stakeholders are normally free to create Tasking Manager projects wherever it’s useful for their operations, with best practice being to consult with affected (OSM) communities whenever feasible.
On the ground data collection presents more potential for harm than remote mapping. In all circumstances, HOT will only engage in or support field mapping if people from all of the potentially affected communities are consulted and give informed consent (and we strongly prefer that the activities be owned and carried out by local communities as far as possible). Even in peaceful situations, context and risk analysis must be done to ensure the potential for negative data use is mitigated. In zones with acute, emergent, or high-intensity active conflict, a very different set of considerations apply. Humanitarian principles and the protection of affected populations take center stage and precedence. This means that our current thinking around this can be summarized as follows:
1. Within zones of active conflict, by default, HOT does not conduct or support additional data creation and mapping—either digitization via the Tasking Manager, or field data collection.
In these situations, humanitarian principles are leading: humanity, neutrality, impartiality, and independence. This often complements HOT’s community driven approach, but can at times also be at odds.
We can consider exceptions to the default position of non-participation in conflict contexts if there is a direct request to do so from trusted humanitarian partners with accurate, reliable, up to date information. However, HOT will not engage until we are satisfied that our context analysis and assessment of potential harm to people is complete and of high-quality (which, in situations where we do not have the relevant expertise, involves consulting with external experts without a stake or interest in the project). Indirect harm, including the risk of compromising humanitarian principles by taking sides, must also be taken into account.
We will refrain from mapping if we can’t be relatively certain, based on well-verified information and diligent analysis that takes into account the knowledge and opinions of the people affected and obtains their informed consent, it can’t be used to identify, target, or otherwise harm people, and there is a clear, compelling, urgent need for the data which is strongly likely to save lives .
We generally support access to, and use of, OpenStreetMap data by humanitarian partners. HOT continues to provide datasets on the Humanitarian Data Exchange (HDX) for Yemen, South Sudan, etc.
In some cases, local communities may request that mapping be suspended during a conflict. This is the case in Ukraine, where a large number of OSM contributors and other community members indicated that they had reason to believe that OSM data was being used by military actors to harm the population. Based on this community request, HOT has decided not to participate in any mapping in Ukraine until further notice.
2. HOT does support humanitarian action for safe migration and displacement in areas and countries neighboring a conflict zone where there is a meaningful partnership with the community and humanitarian partners and a clear path to impact.
In the recent past, HOT has supported major humanitarian response operations around the Rohingya crisis in Bangladesh & Myanmar, Syrian civil war, the conflict-affected areas in the eastern part of the Democratic Republic of the Congo, and the South Sudan civil war. For the latter two, HOT has implemented projects in Turkey with focus on urban refugees in Istanbul and the north-western area of Uganda, and western Uganda.
Projects adjacent to and/or related to conflict necessarily involve vulnerable people and situations where it is hard to be sure what might happen, and therefore require an elevated level of due diligence in context and impact analysis. However, provided they do not involve mapping within the active conflict zone there is not a default assumption that HOT will not participate in them.
3. The humanitarian principle of impartiality is important, which means providing assistance on the basis of the greatest unmet need and potential to save lives and alleviate suffering, not media coverage, donor/funder interest, or partner requests. Before considering requests and beginning the process of risk assessment, we question whether the needs are urgent and unmet. In high profile crises there are often multiple actors providing assistance similar to what HOT might provide. In those cases the needs are not necessarily unmet, which strengthens the default position of non-participation.
There are several community initiatives in countries hosting Ukrainian refugees, and HOT has offered to support such initiatives. On the other hand, we have received at least one request from a community that is not currently within the active conflict zone, but there is credible speculation that the Ukraine conflict will spread to that area; in this case we have regretfully declined to assist. Again, the protection of populations from direct harm resulting from our activities takes precedence.
The OSM Ukraine community is also identifying potentially harmful data and edits, and has requested that the OpenStreetMap Foundation (OSMF) ban users who appear to be persisting in creating data relevant to the military situation. HOT is not involved in either of those issues, but is ready to assist and advise if necessary.
We have added additional language data to our HDX exports for Ukraine based on partner requests. All exports include fields for
name (most commonly used name, usually Ukrainian),
name:uk (Ukrainian name),
name:ru (Russian name), and
name:en (transliterated & translated name to English, if available).
In principle, any activities from HOT around active emergent conflict such as Ukraine will primarily be around trying to bring together (OSM) community needs and requests with the priorities of humanitarian partners. We are monitoring conversations and opportunities in this space.
From here on out, we’re translating our thinking on this topic into concrete guidance and policy. As always, HOT’s leadership, staff, communities, board, and voting members are eager to hear what people think! The conversation is far from over. HOT is currently engaged in a project to build a stronger data protection policy and toolkit, including a much deeper dive into the issues surrounding mapping and data in humanitarian and conflict settings.
 For example, if a humanitarian agency working with a particular government requests support from HOT, it is not sufficient to accept the context analysis from that agency and government, we must also understand the perspective of people on the other side of the conflict to ensure no one is harmed, and therefore consult with people knowledgeable about that context who aren’t specifically motivated to consider the project harmless.
 It’s not actually possible to reduce risk of harm to zero. There is always some residual risk to people. Therefore it is only ethically permissible to even contemplate mapping if there is a well-established case for a benefit that unambiguously outweighs the residual risk.