Symposium on Visual Investigations: Best Practice in Collaboration and Community-led Approaches – Steel, Smoke and Students – Verifying an Industrial Pollution Dispute in Syokimau, Kenya

[Fredah Kanampiu is a journalist in Kenya and a cohort member of WITNESS’ Fortifying Community Truth (FCT) East Africa cohort. 

Mwara Gichanga is a Communications & Engagement Coordinator for WITNESS and part of the Fortifying Community Truth project, led by the Training & Partnerships team. 

Skye Jacob Krabbendam is a student and Seppe Duffhues, Amina Hindawi, Kieran O’Kelly, Suzy-Anne Beekman, and Stella Poulain are recent graduates from various bachelors and masters degrees at Utrecht University. They completed the Open Source Global Justice Investigations Lab with either WITNESS on environmental investigations in Kenya and political violence in East Africa, or with Free Press Unlimited on a confidential project regarding social media disinformation that targets journalists.]

This piece is part of an ongoing exchange between WITNESS and the journalists  trained through Fortifying Community Truth (FCT) – a project that pairs community-rooted reporters with the open-source investigative skills and technical partnerships needed to build  cases that contribute to a justice and accountability pipeline. Below, Kanampiu and Gichanga  describe one such collaboration: an investigation into an industrial pollution dispute in  Syokimau, Kenya. This was done in collaboration with students from the Utrecht University  Open-Source Global Justice Investigation Lab and sketches the line where community  journalism and technical partnership meet. 

Fortifying Community Truth: The East Africa Chapter

Fortifying Community Truth (FCT) began in 2024 in West and Central Africa, built on  the premise that community truths need active defenders and practitioners equipped to push  back when audiovisual evidence is discredited or quietly erased. The project takes direct aim  at a familiar pattern: local journalists sidelined in favour of “outside experts”, even when it is  their proximity and lived experience that give a story its depth. WITNESS equips these  journalists with open-source verification tools and audiovisual technique, and in doing so  backs community-based reporting and verification as legitimate authority in its own right building toward an information ecosystem that is more inclusive, credible, and shaped by the  people closest to the truth. 

The FCT project was expanded into East Africa, bringing in journalists from Kenya, Uganda, Tanzania, Democratic Republic of Congo, and Somalia  – changemakers already embedded in their communities and invested in the future. Training  covered open-source visual verification, archiving, and storytelling, paired with access to  expert advisors from across the continent and beyond. Kanampiu’s case is one example of that  model at work: a journalist’s proximity and trust paired with a student lab’s time and technical range, each unable to build the record alone.

A Community, A Factory, A Question of Proof 

By Fredah Kanampiu 

In Syokimau, on Nairobi’s eastern edge, residents have spent years arguing whether a steel mill moved in on top of their homes or whether their homes went up around the mill. Endmor Steel Millers has denied wrongdoing; NEMA, Kenya’s environmental regulator, has opened and closed inquiries. In between, sit residents describing headaches, soot-covered windowsills, and nights too loud to sleep. My reporting began where most community-rooted investigations do; with people who had already been telling this story for years, with no shortage of documents, court findings, and public complaints – but no single place holding it all together. 

I began by tracing the existing records. One of the earliest I found was a July 2020  Citizen TV exposé, ‘We Can’t Breathe’, which reported on air pollution concerns in  Syokimau. The trail continued through court records, including a 2021 environmental case  brought by the Syokimau Residents Association and others against NEMA and Endmor Steel Millers. The case resulted in a 2022 ruling and a 2023 Environment and Land Court judgment concerning the residents’ request for NEMA to conduct emissions and noise tests at the Endmor plant. 

The dispute is documented outside the courts too: in complaints shared on X, such as a July 2023 post by a Syokimau resident, who said she lived about 500 meters from Endmor and experienced particularly severe air pollution at night; or in a 2025 National Assembly Record concerning complaints about air and noise pollution from Endmor; or another residents’ case filed in 2024 that remains ongoing. 

Building the Record 

Over several months, I gathered court and tribunal records, company filings, and conducted interviews with residents, legal representatives and environmental scientists. The investigation was supported by a partnership with the Global Justice Investigations Lab at Utrecht University. Students working remotely ran geolocation on residents’ social media complaints, mapped wind corridors and airsheds to trace how pollution travelled, and built an interactive timeline cross-referencing housing, industry, legal, health, and media events back to 1954. The students asked me about context: only someone on the ground could say which estate photos were taken in, or whether a complaint had circulated locally before it reached a journalist. In return, I asked the students about methodology: how to read wind-direction data against a residential radius or how to build a source log for a partner organisation. That exchange, more than any single tool, made the investigation.

Continuing the Record 

By the Utrecht University Open-Source Global Justice Investigation Lab

As students, it was our job to understand how corporations in the area affected residents through pollution, land use, and poor residential planning. We were given several research questions, the most fundamental one being: had the industries or the residential areas been there first? 

Satellite imagery and in-person scoping confirmed that land use planning and zoning with poorly implemented physical planning meant there was a lack of adherence to spatial plans and an encroachment of construction, favouring noise and air pollution due to not respecting distancing with industries. Rapid development projects created infrastructure pressure, arriving without the corresponding upgrades regarding sewage, drainage, and water supply, leading for example to illegal waste disposal. We were also able  to geolocate and thus verify several documented complaints that local people had uploaded to  social media that included video and photo evidence of fires and noise pollution. By comparing landmarks and other visual cues in their uploaded evidence to satellite imagery of  the involved companies, we confirmed that the complaints pertained to the companies the  case study focused on. 

For another investigation into electoral violence in an East African state, we worked closely with local journalists who provided relevant sources, such as livestreams and other social media posts. Our job was then to geolocate and verify these sources using OSI methods. 

Outside of the WITNESS collaborations, other students of the Lab worked on a confidential project with Free Press Unlimited (FPU). While not all details of the project are available to disclose, the aim was to look at disinformation surrounding journalists reporting within modern armed conflict. Social media  was the main outlet for this disinformation and became the main field in which the work was done. Hours of scouring through various accounts and posts to see if the sources of such  information could be identified, or if the information itself could be disputed. Tools used included reverse image searching to find the source of the data, as well as geolocating images and videos of individuals appearing to act as fighters or direct participants in hostilities, to verify the presence (or absence) of journalists at specific locations and whether there was other publicly available information that would make their presence at that location  impossible, an ‘alibi’ of sorts. 

Further research showed that there were specific profiles on sites such as X whose primary aim appeared to be discrediting journalists by claiming that they were linked to a  party to a conflict rather than acting as an independent journalist. Oftentimes, the information  used to support this claim came directly from the opposing party to the conflict. While this in  itself does not automatically discredit the claim, it allowed those working on the project to  explore further the links between various accounts and individuals operating within the conflict, using the Maltego tool to begin developing a map of connections between these  accounts and whether there was a coordinated disinformation effort between them. These investigations have shown us how OSI and local knowledge can reinforce one another.

Collaboration 

Only collaboration with journalists working directly in Kenya could have given us the  full picture of the situation in the region. For instance, the lab team was reading Environmental Impact Assessments, documents required by NEMA before building an industry or residential area in Kenya. Amongst these documents, almost all mentioned having asked residents their opinions regarding future constructions via questionnaires. The lawyer representing the residents, however, raised concerns about how the questionnaires had been handled. 

Indeed, media and journalism were the most important tools to establish concrete evidence and enable efficient accountability mechanisms. Journalists’ investigative qualities  allow them to gather evidence and concretely place responsibility. Journalism and media are  community- driven and oriented means that require our attention, participation, support, and  protection. Collaboration between on-terrain journalists and our group of students online  proved how international collaboration is extremely beneficial to investigations such as this  one. Combining the journalists’ expertise with the students’ fresh perspective granted a new  ability for the use of open-source tools for legal and investigative purposes. 

Investigative Justice for Transparency 

In open-source research and investigations, the legal system serves as a ‘standard’ to hold the information you find against. It is the intermediary between doing the investigation and seeking accountability. From constitutional law to executive directives, climate change and protection of the environment is very much addressed in Kenya on paper; these laws cover construction regulations, urban planning, and air and noise pollution. However, gaps in enforcement can hinder implementation, as residential developments ignore stop-orders and bypass compliance regulations. The penalties for environmental offenses can be low and the judicial process equally slow, which may affect enforcement of environmental regulations.

From a legal perspective, the value of OSI investigations is therefore manifold. First, it helps local communities and researchers to collect data and evidence on documented injustices and build a (legal) case for accountability. Moreover, it helps to identify the kind of injustices that have taken place. In the case of environmental pollution in Kenya, the investigations helped the local community to document specifically what kind  of environmental pollution had taken place and identify the relevant laws and regulations. One  example included the damage done by the spread of cement dust over neighbouring  residential areas, combining the study of wind corridors, residences in the area, proximity to  the companies, and health studies done on patients living near them. In our case on electoral  violence, the investigations helped document human rights violations and electoral fraud that took place in the country during elections. This was especially valuable as the government routinely shut down the internet before, during and after the election period.  During those times, the students of the lab could still gain and analyze information, somewhat  “surpassing” the shutdown, aiding the journalists with geolocations almost in real-time.

A second value of OSI investigations is that it helps to enhance transparency and fight disinformation. With the growing use of social media and various AI tools, disinformation is rampant. OSI investigations then help to bring the truth to the people – “to help defend community truths and challenge efforts to discredit or erase crucial evidence”. For example, given journalists were being targeted and killed, an initial reflex of  some of the students working on the FPU project was to assess the lawfulness of this action against the Geneva Conventions, i.e. whether the journalists could be considered as fighters or direct participants in hostilities at the time of the killing. This legal perspective could contribute to government accountability given the responsibilities of any party to a conflict adherent to the Geneva Convention. Moreover, open-source research also benefits from looking beyond legality, with a more journalistic, investigative approach that allows the data to be assessed beyond whether actions were lawful and to question the sources of the data and the validity of the data itself. Once the project moved beyond legality and more towards linking various sources of disinformation, the output became more useful. 

Open source investigations are an important tool for legal communities to fight against the injustices committed by large perpetrators – whether it be companies, governments or large social groups. 

Conclusion 

This collaborative project allowed both the Open Source Global Justice Investigations Lab students, WITNESS, journalists and experts to gain specific insight into the environmental case in Kenya. It shows how open-source investigation, paired with community reporting, builds a dataset of verifiable proof that cannot be easily ignored or discredited. The Syokimau investigation began with a documentary and a set of legal cases against the companies involved: the collaboration that followed helped strengthen the legal arguments, gave greater visibility to the residents of Syokimau and Athi River, and demonstrated how valuable international collaboration can be in investigating cases like this one.

Photo attribution: Pawel Czerwinski on Unsplash