Key Takeaways
Treat public digital traces as evidence, but make every inference checkable.
A new kind of investigation. Open-source investigation means examining material available online, such as social-media posts, videos, satellite maps, and public records, to establish what happened. Eliot Higgins began as an office worker following the Arab Spring, not as a spy or trained journalist. Watching a shaky video from Libya, he realized that its background might contain more reliable clues than the speaker’s claims.
Evidence belongs in public. Bellingcat grew from that insight into a collective that investigates conflict, crime, and disinformation, then shows its sources and reasoning. Its ambition is not to replace reporters, experts, or investigators, but to let ordinary people inspect evidence for themselves. The internet offers clues, not automatic truth: every image can be mislabeled, selectively posted, or misunderstood.
The central innovation is not simply that more people can search online. It is that a claim can be accompanied by an inspectable path from source to conclusion. That resembles reproducible science, where confidence comes partly from letting others examine the method. Yet public availability is not the same as representativeness. Online records overrepresent people with access, incentives, and an audience. The strongest citizen investigation therefore makes two things visible: what the evidence supports, and what the available evidence cannot establish.
Turn an online video into a map by matching independent landmarks.
Geolocation makes images useful. Geolocation is the process of identifying where a photo or video was recorded. Higgins first tried it with a Libyan rebel video said to show fighting in Brega. He sketched the turns in the road, compared the street pattern with satellite imagery, and checked whether background features matched. The result placed the footage in Brega’s eastern residential district, clarifying how opposing sides could both claim parts of the town.
Match more than one detail. A road curve alone could be coincidence. A road curve, junctions, buildings, and open space all aligning make a stronger case. The same approach identified footage from Tiji and later helped locate images of attacks in Syria. Viewers need not understand the language or trust the person filming to test what the setting reveals.
This is a practical lesson in visual reasoning: reduce a complicated scene to features that can be compared across representations. The method also resembles map matching in robotics, where several landmarks help locate a moving observer. Its strength is cumulative confirmation. Its vulnerability is that maps and imagery can be outdated, incomplete, or aligned to the wrong moment. A responsible geolocation therefore checks the date of the underlying map and seeks independent landmarks rather than treating a single apparent match as a verdict.
Build a case from converging clues, not one spectacular image.
Follow the evidence over time. After Malaysia Airlines Flight MH17 was destroyed over eastern Ukraine, Bellingcat volunteers traced a Buk missile launcher through photos and videos. They matched roads and buildings, estimated image times from shadows, and assembled sightings from Donetsk, Zuhres, Torez, and Snizhne into a route for the day of the attack. One later image showed the launcher returning toward Russia with a missile missing.
Different clues answer different questions. A Finnish military specialist examined damage to the aircraft and concluded that it was consistent with a surface-to-air missile. Investigators also matched distinctive wear on the launcher’s side skirt across images, then traced the vehicle to a Russian military brigade. No single clue established the whole account; the value came from pieces that fit together across time, place, and source.
The investigation models a form of cumulative inference: each clue narrows the plausible story, while the combined pattern becomes harder to explain away. This is more robust than a dramatic image because it exposes the links between claims. The caveat is that several clues may not be truly independent. If multiple outlets repeat one original source, their agreement adds little. Strong analysis therefore asks not only how many clues exist, but whether they arise from distinct observations and methods.
Answer propaganda with a public audit trail, not a louder counterclaim.
Contradictions can be tested. After MH17, Russian officials offered competing explanations and presented a video they said placed a Buk launcher in Ukrainian-held territory. Online investigators compared the scene with traffic-camera previews, street imagery, and photos taken at the location. The supposed address on a billboard was absent from the real sign. The visual evidence showed the video had been filmed in Luhansk, in separatist-held territory.
Repetition is not corroboration. The Counterfactual Community is Higgins’s term for a loose network that circulates claims designed to undermine established evidence. Kate Starbird’s study of White Helmet accusations found that a small cluster of websites repeatedly republished the same material. This false triangulation makes a recycled assertion look independently confirmed. Linking to evidence and showing how it was checked gives readers a way to distinguish repetition from verification.
A public audit trail does something more durable than winning an argument in the moment: it lets later readers revisit the underlying evidence. Research on rumor and network communication helps explain why repeated claims can feel convincing even when they have one source. Still, transparency does not guarantee persuasion. People may reject evidence because it threatens a group identity or political commitment. Its value is also institutional: it preserves a record that journalists, researchers, and courts can inspect without relying on the loudest account.
Let findings outrank credentials, and recruit expertise where evidence demands.
Useful investigators come from unlikely places. The early Bellingcat network included a retired history teacher, an office worker, a bank employee, a student, and a former Stasi analyst with 3D skills. During the MH17 inquiry, student Iggy Ostanin searched Russian-language social media and dashcam footage, linking a Buk seen in Ukraine to a Russian military convoy. Other contributors brought language skills, military knowledge, or visual analysis.
Amateurs still need experts. Higgins’s early work on Syrian munitions improved when arms specialists examined his findings. Bellingcat also collaborated with journalists on the ground and shared evidence with official investigators. The principle is not that formal expertise is irrelevant. It is that credentials do not substitute for demonstrable work, and that complex conclusions often require several kinds of knowledge.
The model resembles distributed cognition: a group can solve a problem that no single member could handle, because different people notice different things. It also challenges institutions that confuse status with reliability. The useful counterpoint is that openness alone does not erase skill gaps. Military systems, chemistry, or forensic audio require domain knowledge, and a confident amateur can misread them. The productive arrangement is porous but disciplined: invite diverse contributors, then subject each claim to methods and expertise appropriate to the question.
Publish the reasoning and limits, so skeptics can reproduce the work.
Show how the conclusion was reached. Bellingcat’s practice developed in reaction to journalism built on confidential sources that readers could not assess. Its reports link to source material, explain steps, and mark uncertainty. When Higgins examined Syrian footage, for example, he distinguished what a video showed from what activists claimed it meant. He also sought confirmation from arms experts before treating a weapon identification as solid.
Skepticism needs standards. The failed Reddit search for suspects after the Boston Marathon bombing showed what happens when crowds identify people before verifying evidence. Bellingcat’s alternative is to treat an online claim as a hypothesis, test it against other sources, and avoid speculation. Reproducibility does not make an investigation infallible, but it gives others a way to find errors and correct the record.
Transparency turns criticism into a possible quality-control mechanism. In science, peer review and replication do not prevent every mistake, but they make claims easier to challenge. The same principle matters in public investigations, where an error can damage a person’s safety or reputation. A limitation remains: the most detailed audit trail can overwhelm readers or create a false impression of certainty. Good reporting therefore pairs documentation with plain-language summaries and clearly distinguishes observed facts from inference.
Investigate serious wrongdoing without exposing uninvolved people to needless harm.
Accountability has boundaries. Bellingcat’s investigators set out to identify people linked to crimes, not to name everyone present at a political rally. After a violent assault at Charlottesville, Aric Toler matched a suspect’s distinctive features across photographs and connected him to public social-media profiles. The suspect was later convicted. The team judged the case differently from identifying ordinary demonstrators whose conduct was not criminal.
Protect victims and witnesses. When publishing might alert suspects or endanger people, Bellingcat has sometimes shared information with police before publication. It has also withheld identities of low-ranking soldiers not directly linked to a crime, and referred witnesses to the MH17 investigative team rather than turning their accounts into scoops. The guiding test is whether publication advances accountability without creating avoidable risk.
This is a form of harm minimization, familiar in research ethics and responsible journalism. Digital traces can make identification startlingly easy, but ease is not justification. The Charlottesville example shows a defensible public interest when evidence links a person to a serious assault. The harder cases involve ambiguous conduct, political expression, and mistaken identity. A clear threshold, careful corroboration, and attention to who might be endangered are essential safeguards, especially because online exposure can spread far beyond an investigator’s intended audience.
Archive evidence now, because platforms can erase tomorrow's historical record.
Digital evidence is fragile. Videos documenting Syria’s war disappeared when platforms removed content or services closed. Syrian Archive preserved millions of files, while verifying only a fraction, and Bellingcat found that collecting material was not enough. Investigators needed to record where it came from, how it was handled, and whether it could withstand legal scrutiny. This documented history of handling is called chain of custody.
Preservation can serve justice. In Libya, Bellingcat geolocated videos of executions by Mahmoud al-Werfalli, including one image sequence whose time could be estimated from shadows. The International Criminal Court’s arrest warrant relied largely on social-media material. Such records may matter years after an event, when witnesses are unavailable or authorities have rewritten the story. Preserving context and provenance helps turn online fragments into evidence that can be responsibly assessed.
An archive is not merely a warehouse; it is civic infrastructure for future accountability. Digital records can corroborate testimony, identify patterns, or preserve evidence when institutions fail. Legal use raises a demanding question: can investigators demonstrate that a file is authentic and unchanged? Documentation helps, but platforms control access, formats, and removal policies. Preservation also has ethical costs, since storing graphic material or identifying details can endanger people. Durable archives need security, governance, and rules for responsible access.
Treat extremist humor as camouflage when online performance starts organizing violence.
Jokes can carry a serious program. Bellingcat contributor Robert Evans studied online far-right forums where memes and ironic language helped users approach extremist ideas while denying responsibility for them. After the Christchurch mosque attacks, the killer’s manifesto included deliberately absurd claims intended to distract journalists and entertain forum users. Evans explained how that bait could amplify the attacker’s message if news outlets repeated it uncritically.
Follow behavior, not spectacle. At Charlottesville, investigators identified a man filmed beating a counter-protester by matching details across images and social profiles. More broadly, Bellingcat documented how online subcultures could normalize threats and celebrate real-world violence. The work does not claim that games, memes, or anonymous forums make users violent. It shows why investigators should examine the communities and incentives surrounding an attack, not only the perpetrator’s self-authored narrative.
The important distinction is between content that shocks and content that organizes. A manifesto may be crafted as much for an online audience as for investigators, making amplification part of the attacker’s strategy. Digital anthropology helps here: slang, recurring jokes, and community status can reveal norms invisible to an outsider. But subculture analysis can also overreach, turning association into culpability. Careful work separates evidence of a person’s actions from broad claims about everyone who visits a platform.
Use AI to sort vast archives, but reserve conclusions for accountable humans.
Machines can find patterns at scale. Syrian Archive held about 3.5 million pieces of digital content, far more than a small research team could review manually. Its experiments trained machine-learning systems to flag footage that might show cluster munitions, leaving people to verify the results. Similar tools could search for vehicles, weapons, or repeated versions of the same video.
Synthetic media raises a different problem. Deepfakes use artificial intelligence to create or alter convincing images, audio, or video. They can deceive viewers, but the fear of fabrication can also be used to dismiss authentic evidence. Bellingcat’s answer is to check a clip’s origin, surrounding context, and relationship to other records. Automated tools can narrow the search; they cannot by themselves establish who acted, why, or whether a crime occurred.
The strongest use of AI here is triage, not judgment. A classifier can surface likely matches in a huge archive, helping scarce human attention focus on promising material. Yet training data can encode blind spots, and adversaries can adapt to detection methods. There is also a dual-use problem: tools that help document abuses may help states monitor populations. Human review, transparent uncertainty, and limits on deployment are not nostalgic obstacles to innovation. They are what make investigative technology accountable.
Analysis
We Are Bellingcat is part memoir, part history of a new investigative practice, and part argument for public accountability. Its narrative begins with Higgins’s improvised geolocation of Libyan footage, then follows a transition from solitary hobby to international collective. The cases provide the book’s structure: Syrian weapons and chemical attacks demonstrate how online fragments can illuminate a sealed conflict; MH17 shows how distributed volunteers can build a timeline and test competing claims; the Skripal investigation demonstrates how public records, leaked databases, and specialist collaboration can identify alleged state operatives.
The deeper claim is institutional. Evidence is not trustworthy merely because it comes from a government, a newsroom, or a crowd. Trust grows when the route from source to conclusion is visible and open to challenge. Bellingcat therefore treats publication as both reporting and method: findings are useful not only for what they say, but because readers can inspect how they were reached. This model makes expertise more accessible without pretending that expertise is unnecessary.
The book is strongest when it shows the mechanics of verification and the people who contribute distinct skills. Its accounts also reveal limits. Online evidence is unevenly distributed, often selected by people with political aims, and vulnerable to deletion or manipulation. Open methods do not automatically eliminate bias, and repeated clues may trace back to a single source. The book’s own emphasis on corroboration and uncertainty addresses these risks, though readers should remember that a transparent argument can still be wrong.
Bellingcat’s work also occupies a difficult space between journalism, activism, and law enforcement. Publishing can inform the public, preserve evidence, or help prosecutors, but these purposes can conflict with source protection and privacy. The book’s ethical examples matter because technical capability is not a sufficient justification for identification. Its lasting contribution is a civic model: train more people to examine claims, preserve evidence with care, and make powerful actors answerable to facts that others can see.
Review Summary
We Are Bellingcat receives mostly positive reviews, praised for its fascinating insights into open-source investigations and the organization's achievements. Readers appreciate the book's engaging narrative and its relevance in combating disinformation. Some criticize the repetitive details of investigations and perceive a lack of self-reflection from the author. The book is lauded for its accessibility and importance in understanding modern investigative journalism. Many reviewers express admiration for Bellingcat's work in exposing international crimes and promoting accountability, though a few question the organization's impartiality and funding sources.
People Also Read
Glossary
open-source investigation
Public evidence analysisA method of investigating events by finding and analyzing material available online, including videos, photographs, satellite imagery, social-media posts, and public records. Investigators test what those sources can establish, compare them with other evidence, and document their reasoning so that others can inspect or reproduce the work.
geolocation
Finding an image's locationThe process of determining where a photograph or video was made. Investigators compare visible features, such as roads, buildings, terrain, and signs, with maps or other imagery. A strong geolocation relies on multiple matching details, rather than a single resemblance, and accounts for changes to the location over time.
Counterfactual Community
Network promoting evidence denialHiggins’s term for a loose online network of activists, commentators, conspiracy theorists, and state-aligned media that promotes claims designed to undermine evidence. Its members may have different political positions, but often circulate similar narratives, repeat material across connected outlets, and treat scrutiny as proof of a hidden conspiracy.
false triangulation
Repetition mistaken for corroborationThe misleading impression that a claim has been independently confirmed because it appears in several outlets. In reality, those outlets may all be repeating the same original assertion. Investigators need to trace sources and determine whether apparent agreement comes from genuinely separate evidence or from recycled material.
chain of custody
Record of evidence handlingA documented account of how evidence was collected, stored, transferred, and analyzed. For digital material, it can include where a file was found, who preserved it, and whether it was altered. A clear chain of custody helps others assess authenticity and may be necessary for evidence to be considered in court.
vicarious trauma
Trauma from repeated exposurePsychological harm that can arise from repeatedly encountering distressing material about events experienced by others. Online investigators may watch graphic footage from conflict or crime scenes while working remotely, but the nervous system can still react strongly. Breaks, peer support, careful assignment practices, and limits on unnecessary viewing can reduce harm.
deepfake
AI-generated or altered mediaSynthetic or manipulated media created with artificial intelligence, including fabricated images, altered video, or generated speech that imitates a real person. Such material can deceive audiences or damage reputations. Verification therefore considers not only visual or audio clues, but also a file’s source, context, and consistency with other evidence.
FAQ
What's We Are Bellingcat about?
- Focus on Open-Source Investigation: The book explores the rise of Bellingcat, a collective that uses open-source intelligence (OSINT) to investigate global events, such as war crimes and disinformation campaigns.
- Real-World Applications: It details investigations like the Skripal poisoning and the MH17 downing, showing how ordinary citizens can uncover truths using publicly available information.
- Empowerment of Individuals: The narrative emphasizes that anyone can engage in investigative work, challenging traditional journalism and intelligence gathering.
Why should I read We Are Bellingcat?
- Insight into Modern Investigations: The book offers a unique perspective on using digital tools and social media for investigative journalism, relevant for those interested in current affairs.
- Inspiration for Activism: It serves as a motivational guide for holding power accountable, demonstrating that collective action can lead to significant change.
- Understanding Disinformation: Readers gain insights into tactics used by state and non-state actors to manipulate information, crucial in today's media landscape.
What are the key takeaways of We Are Bellingcat?
- Importance of Transparency: Open-source investigation relies on transparency and verifiable evidence, contrasting with traditional journalism's reliance on anonymous sources.
- Community Collaboration: The book highlights the power of collective effort, showing how diverse individuals can uncover truths that would otherwise remain hidden.
- Digital Tools for Justice: It illustrates how technology can document and expose human rights abuses, empowering citizens to demand accountability.
What are the best quotes from We Are Bellingcat and what do they mean?
- “We insist on evidence.”: This quote encapsulates Bellingcat's commitment to factual accuracy and the importance of substantiating claims with verifiable data.
- “Bellingcat is something that has never been before: an intelligence agency for the people.”: It reflects the book's core message that investigative work is accessible to anyone with information.
- “The response to information chaos is transparency.”: This emphasizes the need for clarity and openness in an age of misinformation, advocating for a methodical approach to uncovering the truth.
How did Bellingcat start its investigations?
- Origins in the Arab Spring: Eliot Higgins began investigating events during the Arab Spring, using social media and online videos to gather information.
- Development of Techniques: Higgins and others developed methods like geolocation and OSINT, foundational to Bellingcat's approach.
- Community Formation: Initial efforts attracted a community of like-minded individuals, leading to Bellingcat's establishment as a formal entity.
What is OSINT and how is it used in We Are Bellingcat?
- Definition of OSINT: Open-source intelligence involves collecting and analyzing information from publicly available sources, such as social media and satellite imagery.
- Application in Investigations: Bellingcat uses OSINT to verify claims, geolocate events, and uncover evidence of wrongdoing, demonstrating its effectiveness in modern journalism.
- Empowerment through OSINT: The book shows how OSINT allows ordinary citizens to participate in investigations, challenging traditional power structures.
What are some notable investigations detailed in We Are Bellingcat?
- Skripal Poisoning Case: Bellingcat identified suspects involved in the poisoning of Sergei Skripal, showcasing OSINT's power in high-profile cases.
- MH17 Downing: The investigation traced the missile's origin using social media and satellite imagery, highlighting Bellingcat's methods.
- Chemical Attacks in Syria: Bellingcat documented these events using open-source evidence, attributing attacks to the Assad regime.
How does We Are Bellingcat address the issue of disinformation?
- Exposing False Narratives: Bellingcat combats disinformation by verifying facts and providing clear evidence, countering misleading claims by state actors.
- Case Studies of Disinformation: The book provides examples of disinformation campaigns, particularly in the Syrian civil war and the MH17 incident.
- Call for Vigilance: It emphasizes the need for critical thinking and skepticism, urging readers to seek out verified sources.
What role does technology play in Bellingcat's investigations?
- Utilization of Digital Tools: Bellingcat leverages technology, such as satellite imagery and social media analytics, to enhance investigative capabilities.
- Innovative Techniques: The book discusses methods like geolocation and metadata analysis, essential in modern investigations.
- Accessibility of Technology: These tools are available to anyone, democratizing the investigation process and empowering citizens to seek the truth.
How does We Are Bellingcat envision the future of journalism?
- Shift Towards Citizen Journalism: The book suggests that journalism will increasingly involve citizen participation, with ordinary people taking on investigative roles.
- Need for Adaptation: Traditional media must adapt to the changing landscape, embracing new technologies and methods to remain relevant.
- Importance of Collaboration: Collaboration between journalists, activists, and the public will be crucial in holding power accountable and ensuring the truth prevails.
How does Bellingcat conduct its investigations?
- Open-Source Research: Bellingcat relies on publicly available information, such as social media posts and satellite imagery, to gather evidence.
- Verification Techniques: They employ rigorous methods to ensure accuracy, including cross-referencing sources and using geolocation.
- Community Engagement: Bellingcat involves the public in investigations, expanding research capabilities and empowering individuals to contribute.
What ethical considerations are discussed in We Are Bellingcat?
- Responsibility to Protect: The book emphasizes protecting vulnerable individuals during investigations, especially in conflict zones.
- Transparency in Methods: Higgins advocates for clear communication of methods and sources, building public trust and credibility.
- Impact on Subjects: The book discusses considering how investigations affect individuals and communities, advocating for a respectful approach.
Download PDF
Download EPUB
.epub digital book format is ideal for reading ebooks on phones, tablets, and e-readers.