Media Manipulation and Bias Detection
Auto-Improving with AI and User Feedback
HonestyMeter - AI powered bias detection
CLICK ANY SECTION TO GIVE FEEDBACK, IMPROVE THE REPORT, SHAPE A FAIRER WORLD!
Haitian Bridge Alliance / immigrant-rights advocates
Caution! Due to inherent human biases, it may seem that reports on articles aligning with our views are crafted by opponents. Conversely, reports about articles that contradict our beliefs might seem to be authored by allies. However, such perceptions are likely to be incorrect. These impressions can be caused by the fact that in both scenarios, articles are subjected to critical evaluation. This report is the product of an AI model that is significantly less biased than human analyses and has been explicitly instructed to strictly maintain 100% neutrality.
Nevertheless, HonestyMeter is in the experimental stage and is continuously improving through user feedback. If the report seems inaccurate, we encourage you to submit feedback , helping us enhance the accuracy and reliability of HonestyMeter and contributing to media transparency.
Using emotionally charged language to provoke feelings rather than focusing strictly on neutral, verifiable description.
Examples include: 1) “Damas’ death is part of a deeply troubling pattern.” 2) “Far too many immigrants—including Haitian nationals—have died in ICE custody.” 3) “The death of Emmanuel Damas is a devastating reminder that our immigration detention system is failing the most basic standard of human dignity. People seeking safety should not die from untreated medical conditions while in government custody.” These statements are advocacy-oriented and framed to elicit moral outrage and sympathy. They go beyond describing the specific incident and use strong evaluative terms like “deeply troubling”, “far too many”, “devastating reminder”, and “failing the most basic standard of human dignity.” While appropriate for a quoted advocate, they are still emotional framing rather than neutral description.
Replace highly emotive adjectives with more neutral wording while preserving the core concern. For example: change “Damas’ death is part of a deeply troubling pattern” to “HBA says Damas’ death is part of a broader pattern of deaths in ICE detention.”
Qualify broad moral judgments with attribution and, where possible, data. For example: change “our immigration detention system is failing the most basic standard of human dignity” to “HBA argues that the immigration detention system fails to meet basic standards of humane treatment, citing recent deaths in custody.”
Add factual context alongside emotional claims. For example, after “Far too many immigrants—including Haitian nationals—have died in ICE custody,” include comparative figures or official statistics and note any disputes or alternative interpretations from independent experts or ICE.
Reducing a complex issue to a simple, categorical claim without acknowledging nuance or uncertainty.
The quote: “The death of Emmanuel Damas is a devastating reminder that our immigration detention system is failing the most basic standard of human dignity. People seeking safety should not die from untreated medical conditions while in government custody.” This presents the entire immigration detention system as categorically “failing” based on a set of deaths, without exploring variations between facilities, existing medical protocols, or ongoing reforms. It also assumes the cause of death (“untreated medical conditions”) as a systemic failure before any investigation results are presented.
Introduce nuance and attribution. For example: “HBA contends that repeated deaths in detention indicate serious shortcomings in how medical care is provided in some facilities.”
Clarify what is known versus alleged. For example: “HBA says Damas’ death, which they allege was linked to complications from an untreated tooth infection, raises questions about whether medical care in detention meets basic standards.”
Add context about the complexity of the system, such as mentioning that conditions and medical care can vary by facility and that investigations are ongoing or pending.
Presenting numerical claims or patterns without sourcing, context, or acknowledging uncertainty, which can amount to selective use of data.
The article quotes: “Damas’ death is part of a deeply troubling pattern,” and then: “at least 30 to 32 people died in ICE detention in 2025. ‘The crisis appears to be continuing into this year,’ Jozef continued. ‘Multiple deaths have already been reported in the early months of 2026, including four migrants who died while in US immigration custody in the first 10 days alone in 2026.’” These figures are presented solely via HBA, with no indication of the source (ICE data, NGO reports, media tallies) and no baseline for comparison (e.g., total detainee population, historical trends, or whether ICE contests these numbers). Calling it a “crisis” is also a strong evaluative label without independent corroboration in the article.
Explicitly attribute the numbers and provide or reference a source. For example: “According to HBA, citing [ICE statistics / NGO report / media compilation], at least 30 to 32 people died in ICE detention in 2025.”
Add contextual data to avoid cherry-picking. For example, include total detainee numbers, historical death rates, or comparisons to previous years, and note if ICE or independent experts interpret the data differently.
Qualify the term “crisis” as an opinion. For example: “Jozef described the situation as a ‘crisis,’ arguing that the number of deaths indicates systemic problems, a characterization ICE has not publicly endorsed.”
Leaving out relevant context or responses from other key actors, which can tilt perception even if the text is factually accurate.
The article includes ICE’s brief description of Damas as “a criminal illegal alien from Haiti currently facing criminal charges for assault and battery” and the date of death, but it does not include: 1) Any ICE response to the specific allegation that Damas died from complications of an untreated tooth infection or that medical care was inadequate. 2) Any mention of ICE’s stated medical protocols, investigation procedures after in-custody deaths, or whether an internal or external investigation has been opened. 3) Any independent expert or third-party perspective on deaths in immigration detention. As a result, the narrative is largely framed by HBA’s interpretation of events and systemic causes, with ICE’s role limited to a short characterization of the detainee and confirmation of death.
Include ICE’s response to the medical-care allegation, or clearly state that ICE did not respond by deadline. For example: “ICE did not respond to questions about the medical care Damas received before his death” or “ICE said it could not comment pending an investigation.”
Add information on standard procedures after a death in custody (e.g., autopsy, internal review, involvement of independent oversight bodies) and whether such steps are underway in this case.
Incorporate at least one independent source (e.g., medical experts, human-rights monitors, or official oversight reports) to provide context on detention conditions and mortality, rather than relying solely on an advocacy group’s framing.
Using strongly value-laden or categorical terms that implicitly judge one side without qualification.
Within the HBA quotes: - “deeply troubling pattern” - “The crisis appears to be continuing into this year” - “failing the most basic standard of human dignity” These phrases are not neutral descriptions; they embed a strong negative evaluation of ICE and the broader detention system. While they are clearly attributed to HBA, the article does not balance them with equally detailed or contextualized language from ICE or neutral observers.
Maintain the quotes but frame them explicitly as advocacy positions. For example: “Jozef, whose organisation advocates for immigrants’ rights, described what she sees as a ‘deeply troubling pattern’…”
Add neutral paraphrasing alongside the quotes to clarify that these are evaluative judgments. For example: “She argued that the number of deaths indicates systemic failures in detention conditions.”
Balance the evaluative language with more detailed representation of ICE’s position or relevant official findings, if available, so readers can weigh competing interpretations.
- This is an EXPERIMENTAL DEMO version that is not intended to be used for any other purpose than to showcase the technology's potential. We are in the process of developing more sophisticated algorithms to significantly enhance the reliability and consistency of evaluations. Nevertheless, even in its current state, HonestyMeter frequently offers valuable insights that are challenging for humans to detect.