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!
Israel / United States and allies
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.
Use of value-laden or one-sided terms that implicitly endorse one side’s framing.
1) “The IDF (military) has begun an additional wave of strikes on Iranian terror regime targets,” the Israeli military wrote on its official Telegram channel. 2) “Lebanon was drawn into war last week when Iran-backed militant group Hezbollah attacked Israel in response to the killing of Iranian supreme leader Ayatollah Ali Khamenei in US-Israeli strikes.” In (1), the phrase “Iranian terror regime targets” is a highly loaded characterization. It is presented as a quote from the Israeli military, but the article does not explicitly signal that this is partisan language or provide any balancing description from the Iranian side or neutral observers. In (2), “Iran-backed militant group Hezbollah” is a common journalistic formulation, but it still reflects a particular framing (militant vs political party / armed group) without clarifying that this is a contested designation in some contexts.
Make clear that loaded terms are the speaker’s characterization, not the outlet’s voice. For example: “The IDF (military) said it had begun an additional wave of strikes on what it called ‘Iranian terror regime targets’…”
Where possible, add neutral descriptors alongside partisan ones. For example: “Hezbollah, an Iran-backed armed and political group designated as a terrorist organisation by some countries, attacked Israel…”
Avoid adopting any side’s pejorative labels in the reporter’s own narrative voice; keep such language strictly within quotation marks and attribute it clearly.
Leaving out important context that would help readers fully understand events and competing perspectives.
The article lists a sequence of developments but provides almost no background on: - The broader origins and timeline of the conflict (“Middle East war” and “Iran war” are referenced without explanation). - Civilian casualties and humanitarian impact on all sides, beyond a single displacement figure for Lebanon (“nearly 760,000 people had been registered as displaced”). - International law or independent verification of claims (e.g., the killing of four Iranian diplomats, the nature of alleged spying, the legal status of strikes in Beirut and Tehran). - How different actors (e.g., Lebanese government, independent human rights groups, UN bodies) view these events. This omission does not necessarily show deliberate bias, but it limits readers’ ability to critically assess the claims and relative responsibilities of each side.
Add a brief background paragraph summarizing how and when this phase of the conflict began, including key triggering events and prior escalations.
Include at least one sentence on reported civilian casualties and humanitarian conditions in affected areas (Iran, Israel, Lebanon, Gulf states), citing neutral sources such as UN agencies or reputable NGOs.
When reporting serious allegations (e.g., killing of diplomats, espionage arrests), note whether claims have been independently verified and, if available, include responses from the accused side or neutral observers.
Clarify contested terms like “Iran war” by specifying that this is a shorthand for the current confrontation and not a formally declared war, if that is the case.
Presenting information in a way that gives more space or credence to some actors’ narratives than others, without clear justification.
The piece alternates between statements from multiple actors (Israel, Iran, US, Russia, Ukraine, UAE), but the structure and selection of quotes can subtly skew balance: - Israeli and US military actions are described with operational language and their justifications (e.g., “conducted a precise strike targeting key commanders”). - Iran’s actions are often framed via accusations (e.g., “Iran’s intelligence ministry announced the arrests of 30 people accused of spying ‘on behalf of two Persian Gulf countries in the name of the American-Zionist enemy’”) without any external assessment of credibility. - There is no independent or third-party commentary to contextualize or challenge any side’s claims. While both blocs are quoted, the absence of neutral analysis or counterpoints can make official narratives appear more authoritative than they are.
For each major claim by one side, where feasible, include either a response from the other side or a note that the other side has not yet commented.
Add brief input from neutral or expert sources (e.g., regional analysts, UN officials) to contextualize military claims and political statements.
Clarify when information comes solely from one party and has not been independently verified, using consistent language for all sides (e.g., “The Pentagon said…; the figures could not be independently verified”).
Using emotionally charged imagery or descriptions that may influence readers’ feelings more than their understanding.
The caption: “Smoke billows from the site of an Israeli air strike in the southern suburbs of the Lebanese capital Beirut…” accompanied by a war image can evoke strong emotional reactions. On its own this is standard news practice, but when combined with minimal context about casualties, targets, or legality, it can steer readers’ emotional response without fully informing them.
Pair emotive images with concise factual context about what is known and unknown (e.g., targets, casualties, responsibility, and verification status).
Ensure that imagery over time reflects suffering and impact on all affected populations, not predominantly one side, to avoid skewing emotional perception.
Where details are uncertain, explicitly state that information is still being verified rather than letting the image imply more than is known.
Reducing a complex, multi-actor conflict to a simplified narrative that may obscure important nuances.
Phrases such as “Lebanon was drawn into war last week when Iran-backed militant group Hezbollah attacked Israel in response to the killing of Iranian supreme leader Ayatollah Ali Khamenei in US-Israeli strikes” compress a complex chain of events and long-standing tensions into a single cause-and-effect sentence. This can imply a neat linear sequence and unified national positions (e.g., ‘Lebanon’ as a whole) that do not reflect internal divisions or prior hostilities.
Break complex causal chains into multiple sentences that acknowledge uncertainty and multiple contributing factors. For example: “Hostilities expanded to Lebanese territory last week after Hezbollah attacked Israel. Hezbollah said the attack was in response to the killing of Iranian supreme leader Ayatollah Ali Khamenei in US-Israeli strikes. Lebanon’s government has taken [describe stance if known], and the country remains internally divided over Hezbollah’s role.”
Avoid attributing actions of non-state actors to entire countries without clarification (e.g., distinguish between ‘Lebanon’ and ‘Hezbollah’).
Where the causal link is based on one side’s stated justification, make that explicit (e.g., “Hezbollah said it was responding to…”).
- 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.