DeBunKeD
Misinformation awareness and debunking fake news service
Case Study
Monsoon 2024
Work
Desk research, Literature review, Interface design, user interview, design thinking.
Duration
July - September 2024 (10 weeks)
Debunkd is a news aggregator platform similar to Google News. This service will analyze news articles, gather relevant information, and assess data to verify or refute claims.
Our goal is to provide content that guides readers toward verified, de-sensationalized news.
End to end design project, from Research to prototype.
Misinformation has led to a global trust deficit, with only 40% of people trust the news they consume.
The Problem
Why is misinformation a big problem?
An example
How bad can the affect of fake new be?
Misinformation about COVID-19 vaccines has significantly impacted vaccination rates. A study estimated that exposure to antivaccine content on Twitter led approximately 750,000 individuals in the U.S. to refuse vaccination.
Resulting in at least 29,000 additional cases and 430 additional deaths.
The spread of true and false news online
9 Mar 2018
Fake news spreads
10x faster
than true news, especially during political events.
Factors such as
Lack of deliberation
Repeated exposure
to false information make individuals more likely to believe it.
Misinformation can shape public behavior, impacting areas like health and political engagement.
The Goal
We seek to make the user more sensitive and aware of their consumption of misinformation and notice biases and prejudices around them and empower them with the information they need to think critically and rationally about issues.
In the end we want to make users feel confident and safe about the information they consume and restore their trust in journalism and news reporting.
An example
Timeline
Discover
Effects of Fake news:
Some Datapoints
Literature Review 1.1
The spread of true and false news online
9 Mar 2018
The top 1% of false-news spread routinely between
1000-100,000
People
It took the truth about
6x
as long as falsehood
reach
1500
People
Misinformation and disinformation has risen rapidly in rankings to 1st place for the two-year time frame, and the risk is likely to become more acute, during elections.
Preference for reading news and getting current affairs.
YouTube (31%)
WhatsApp (21%)
Twitter/X(10%)
India’s position in
misinformation
Literature Review 1.2
1st risk in India
Misinformation and fake news is considered the biggest risk
2023-2024
India-specific
data points of fake news
Digital News Report indicated that around 63% of Indian respondents had encountered misinformation online.
A study found that 58% of Indian voters believed that misinformation affected their voting decisions
A study during the COVID-19 pandemic revealed that 87% of respondents in India had encountered misinformation related to the virus
An example
Global trends in news consumption and the challenges posed by misinformation
58%
Get news via WhatsApp
The Psychology of Fake News and Sensationalism
Literature review 1.3
How does the mind works?
The psychological reason for misinformation
How does misinformation still exist in this age of information?
Intuitive vs Analytical thinking
Misinformation can shape public behavior, impacting areas like health and political engagement
Illusory truth.
People tend to believe information that they’ve been repeatedly exposed to, even if it’s false. Familiarity breeds perceived truth.
The continued influence effect
Even after misinformation is corrected, people often retain the original false belief, which continues to influence decision-making and memory.
Findings
The brain often pulls out the misinformation (fake news) over the corrective information (debunked news). Since the misinformation is often highly emotional and may appeal to the users worldview better.
“In one study, participants read positive, neutral and negative headlines about the actions of specific people; social judgements about the people featured in the headlines were strongly determined by emotional valence of the headline but unaffected by trustworthiness of the news source.”
Psychologists justified three different models of how misinformation and corrective information are stored.
Literature review 1.4
Theory 1
Competition of Information
Both the memory of misinformation and Corrective information coexist in the brain simultaneously and sometimes we may end up picking the misinformation over the corrective information as its “bigger”.
Theory 2
Layered Encoding
Corrective information is layered onto misinformation like an onion; if the layers are too thin, the corrective information may not be fully processed.
Theory 3
Associative Links
Misinformation and corrective information are separate but linked. Recall of misinformation triggers the corrective info only if their connection is strong enough.
Findings
All three of these models suggest increasing the association between misinformation and corrective information and increasing the psychological impact of corrective information
Focusing on Prebunking and inoculation theory
Literature review 1.5
From last finding
Provide a factual account that includes an alternative explanation for why something happened. For example, if a fire was thought to have been caused by negligence, then providing a causal alternative (‘there is evidence for arson’) is more effective than a retraction (‘there was no negligence’).
Provide a factual account that includes an alternative explanation
Provide a factual account that includes an alternative explanation for why something happened. For example, if a fire was thought to have been caused by negligence, then providing a causal alternative (‘there is evidence for arson’) is more effective than a retraction (‘there was no negligence’).
Point out logical fallacies
Provide a factual account that includes an alternative explanation for why something happened. For example, if a fire was thought to have been caused by negligence, then providing a causal alternative (‘there is evidence for arson’) is more effective than a retraction (‘there was no negligence’).
Corrections are more impactful
Provide a factual account that includes an alternative explanation for why something happened. For example, if a fire was thought to have been caused by negligence, then providing a causal alternative (‘there is evidence for arson’) is more effective than a retraction (‘there was no negligence’).
Do not mention the misinformation more than once.
Conducting One on One Interview
Primary Research
From last finding
Conducted one on one interviews with news readers, psychologists and others involved in the news media industry.
Research Goals
To understand the reasons and pathways through which misinformation spreads and reaches people.
To understand user’s mindset on trusting and not trusting news articles.
To find out existing methods people use to combat misinformation currently.
To understand the reasons and causes of the recent decline in quality of journalism.
To find out why news medias are ineffective in combating misinformation
Insight 1
News from Social Media
Users primarily get news from reposts and shares on social media, shaping their opinions based on these sources.
Insight 2
Missed Corrections
News topics evolve over months, but users often miss updates or corrections as they quickly lose interest in ongoing stories.
Insight 3
Verifying News
Users check multiple sources to verify news but may have skewed perceptions about the credibility of these sources.
Insight 4
Debunking Challenges
Refuting misinformation is difficult, often leading to unproductive arguments that damage relationships.
Insight 5
Limited Awareness of Fact-Checking Sites
Users are largely unaware of debunking sites like Factly or Snopes, and these platforms have poor user experiences for their purpose.
Findings
People rely on social media for news, often missing updates and struggling to verify credibility. Refuting misinformation leads to conflicts, while debunking platforms like Factly or Snopes remain underutilized due to poor user experiences.
Understanding our Market Point of view
Competitor Analysis
Chances to get in business
Direct Competitors - News Aggregator
Fact checking news sources and news sources with fact checking wings.
Pros
Some news sources have different sections for real and fake news, They have some distinction between real and fake news with tags and seals
Some sources are very thorough with their debunking, they seem to consider a lot of factors and and sources to put in a lot of effort to debunk information
Cons
They run on donations, so they seem to be limited in scale of operations
No social media outreach
The misinformation itself is repeated far too many times within the interface, this may lead to Illusory effect
The debunking is far too long and lengthy to be scanned quickly, this leads to disinterest in the application
Snopes, Factcheck.org, Alt news, Factly.
Indirect Competitors - Social media
Social media fact checking programs and policies.
Pros
Action is taken very quickly, The repercussions to spreading misinformation are quick and impactful.
They have highly developed guidelines and parameters to categorize fake news and take appropriate action for it.
There is good collaboration with third party providers who are certified (ICFN)
Cons
Content moderation is unable to parse thorough content in regional languages
There is no redressal mechanism in place once a content is deemed as misinformation
Content moderation is highly circumstantial. Posts are judged on an individual basis without any contextual awareness
Facebook factchecking programs, Twitter/ X Birdwatch, YouTube fact check panels.
Indirect Competitors - Analytical AI
Analytical / AI-ML based debunking services
Pros
Can analyze huge amount of news sources to spot misinformation
Can be scaled easily to track multiple sources of information simultaneously.
AI models can be trained to work across different languages and platforms, enabling debunking services to address misinformation
Service works without human intervention
Cons
Contextual nuances, that are cultural are hard to differentiate by AI models.
Require human overview, to notice mistakes
AI models can have biases that influence the outcome of the news
Newsguard, Logically, Deepware Scanner
Indirect Competitors - Browser EXT
Browser extensions, internet tools
Pros
Highly efficient and scalable AI systems with a focus on real-time analysis.
Content validation right from the source.
Visual insights into how misinformation spreads, aiding in strategic interventions
Great handy tool to check opened articles
Cons
Possible mistakes can be made by the tool as they work solely without human observation
Often bad experience or method of use.
Overall user accessibility becomes a problem for novice users
Focuses on training over direct misinformation identification
Adverif.ai, InVID, Hoaxy
Findings
Analytical AI and browser extensions demonstrate the potential for automated and scalable fact-checking.
Competitors struggle with addressing regional languages and cultural nuances.
Browser extensions and social media platforms often face usability and accessibility issues.
News aggregators collaborate with certified third parties to improve credibility.
Exploring a sustainable business model, such as subscription tiers or partnerships, can ensure scalability.
Define
ERAF and pestel analysis to identify hidden influencing factors
Stakeholder Analysis
Entity Relationship attributes flow(ERAF)
To gain more clarity on how news media interacts with other sections of society, we mapped them out to form a system.
Insight 1
The Sensationalism Feedback Loop in Modern Media
Readers, advertisers, news outlets, and social media form a self-perpetuating cycle that fuels sensationalism. As readers are drawn to sensationalized news, advertising profits increase, prompting news organizations to produce more of the same content. This creates a continuous, negative feedback loop.
Insight 2
The Role of Political Influence and Advertisers in Information Suppression
Political parties and advertisers play a significant role in shaping media narratives, often leading to the suppression of information. This results in biased reporting or the selective amplification of certain stories, contributing to misinformation and the omission of important news.
PESTEL Analysis
Government Censorship Challenges
Legislation like the Telecommunications Bill of 2023 and the Digital Personal Data Protection Act of 2023 may limit the service’s freedom, particularly if it is seen as opposing state interests, potentially leading to legal issues or shutdowns.
Insight 1
Evolving Misinformation Technologies
As AI-generated misinformation and deepfakes become more advanced, debunking services may struggle to keep up with these rapid changes.
Insight 2
Public Distrust and Perceived Bias
Public skepticism toward media and fact-checking, especially among polarized groups, could undermine the service’s effectiveness. If users view it as biased, they may reject its corrections.
Insight 3
Algorithmic Content Limitations
Refuting misinformation is difficult, often leading to unproductive arguments that damage relationships.
Insight 4
Cybersecurity Risks
Handling sensitive information may make the service a target for cyberattacks from those who profit from spreading misinformation.
Insight 5
User Persona 1
Santoshi
28 y.o A Concerned Parent
Santhoshi works as a software engineer at an MNC in Bangalore. She raises a 3 year old kid with his partner and is currently looking for good nurseries around the city. Her and her husband consume a lot of news through social media, a lot of which is focused on parenting, child education and family well-being. They are vary of the news their friends post in social media and want a reliable source of information for news that relate to their kid.
Goals
To stay informed about the latest developments in child education and parenting.
To ensure the information he reads is credible and factual
To raise his child with best practices supported by evidence, not myths or misinformation.
Frustrations
She is overwhelmed by conflicting news articles about parenting, child nutrition and education.
She is concerned about misinformation on media that can’t be easily trusted.
“I really don’t know what to believe or not believe when it come to raising my kid”
User Persona 2
Himavanth
47 y.o Whatsapp warrior
Himavanth runs his boutique mostly alone, spending his time on his phone reading news, browsing social media, and texting his community people in their WhatsApp group. As their lives have diverged, disagreements, especially on politics, have become common. Well-read on current affairs, Himavanth often spots fake news but hesitates to correct his friends, fearing it will lead to arguments. He seeks a calm, effective way to share accurate information without conflict.
Goals
To stay well-informed about general affairs and ensure that the information he holds is accurate.
To find non-confrontational ways to correct misinformation in his WhatsApp group without escalating fights or arguments.
To be confident that the news he shares with others is factually correct.
Frustrations
He finds it difficult to convince his friends that certain news stories are inaccurate without sounding confrontational.
He sometimes questions the reliability of the news he reads, especially when multiple sources provide differing takes.
“I just want to get along with my friends and have mature conversations with them”
User Persona 3
Akshaya
32 y.o apollo tyres, VP
Akshaya is the VP of logistics for her company. She needs to stay updated on topics related to shipping and trade. With the recent trade restrictions happening due to the Ukraine war, she is struggling to stay on top of the current situation there. She constantly checks her phone for updates and monitors the situation carefully as it has vast consequences for the company
Goals
To stay on top of recent developments in her industry and monetize that information to her advantage.
To gain a competitive advantage over her peers by being tuned to news.
Frustrations
It’s difficult to parse out impactful, true information from the deluge of irrelevant and harmful content in the internet.
It’s difficult to parse out impactful,
“I need to be on top of so many things. I wish i could easily stay updated on things i find important easily in my busy schedule”
With So much info and direction, i defined the important stuff
Requirements & constraints
Defining them to get the direction
Requirements
The core principles of pre-bunking (preventing misinformation before it spreads) and debunking (correcting false information) should be reflected in the website’s design
The solution should incorporate feasible and sustainable income streams.
The platform should be easily accessible and integrated with tools that users already rely on, such as WhatsApp, social media apps, and news aggregators.
We need to come up with new workflows that efficiently address the needs of the users to access fake news
Users should be encouraged to gain a broader understanding of the news, beyond just reading articles.
The user must gain information from the product that empower them to spot fake news and argue against it more effectively
The solution must build a strong sense of trust and social proof with the users
Constraints
The solution should promote long-term user adoption, not just serve as a one-time fact-checking tool.
Our solution mustn't rely heavily on advertisements that may force us to compromise on the quality of our content later.
The solution should ensure that fake news is not inadvertently amplified by overemphasizing misinformation.
Misinformation should only be highlighted alongside corrections and verified facts to prevent further spread.
The platform must account for the growing prevalence of AI-generated content and deepfakes in spreading misinformation.
But How might we ?
Hmmmmmmmmmmmmmmmm
…Ensure users get the full picture of a news story as it evolves?
…Make it easier for the users to compare between many sources and reach a mental consensus on news?
…De-sensationalize the news a user might see?
…Make the user be more proactive and analytical about any news they read, on our platform or otherwise?
…Make it easier to have a productive discussion about misinformation?
…Tackle the problem of sharing misinformation?
Ideate
\\brainstormed, curated, categorized ideas using MoSCoW, problem solution matrix, reverse brainstorming and more
Came up with over 24 feature ideas.
We realized that the main reason for the lack of innovation in news debunking platforms is inefficient business models that don’t generate enough revenue and fail to deliver sufficient value to users.
Business Model Canvas
How can this turn into a business
Key Partnership
News channels
Journalists and Media
Data Analysts
Government
Social Media Users
Key Activities
Investigations
Content Databases crosschecks Information
AI and machine learning models to detect fake news
Social Media content creation
Established media outlets.
Key Resources
Investigators
Data analysts
Researchers
Journalists
Content databases and APIs
AI and machine learning models
Value Proposition
Debunking fake news
Reaching people with correct information
Keeping updated as news evolves
Preventing sensationalism and useless news
Spreading awareness about disinformation
Customer Relationship
Social benefits
Neutral medium to the audience
News readers
Fair news source
Journalists
Opportunity
Distribution Channels
Social media
Web news articles
WhatsApp messages
Word of mouth
Ads
Partnerships
Customer Segment
News readers
Industry experts
Investors
Journalists
Debaters
Cost Structure
Cost of Revenue
IT infrastructure
Cloud Computing
General & Admin
Salaries for fact-checkers
content curators
data scientists
Legal Professional
Marketing and outreach campaigns.
Revenue Streams
Donations
VIP Subscriptions
Ads in other social media channels
Data licensing
Freemium
Collaboration
We realized that the main reason for the lack of innovation in news debunking platforms is inefficient business models that don’t generate enough revenue and fail to deliver sufficient value to users. Therefore, we decided to start by creating a Business Model Canvas (BMC) and setting constraints to guide the development of our solution.
Being a news aggregator and not a news publication:
The news aggregator market was valued at 2.15 billion USD in 2023 and has consistently expanded since then. It’s forecasted to rise at a CAGR of 11.2% for the upcoming five years.
there is a huge growing demand for individualized news experiences.
Being a news aggregate enables us to keep our team lean and sharpens our focus on debunking news and curbing misinformation. It also acts as a mechanism for us to incentivize news media to publish non plagiarized, non sensationalized true news.
Free of ads
We realized that most users are cynical of newspapers because of their approach to advertising. We decided to keep the platform free of ads to build trust with them and to avoid falling prey to the cycle of sensationalism on the print media platform.
Monetization
Three unique ways to earn money instead.
Social Media Channels
We can use our content to run associate social media channels (think morning brew) that can promote our platform as debunking news is inherently sensational, and make money through collaborations and sponsorships that are carefully curated by our team to maintain our trustability.
Monetization 1
Professional Debunking Services
Our platform will offer professional debunking services for individuals and businesses. Clients can seek our expertise to navigate complex news scenarios and gain clarity on what is true, empowering them to make informed decisions.
Monetization 2
Subscription/ Premium Features
Website articles would be availbale for all user, and so does Whatsapp debunking , but premium features will include Website extentions and Voice assisted check in with Siri or Gemini.
Monetization 3
Lets explore the solution and how it works with the help of our personas, Santhosh, Himavanth and Himanshi
User flow
Analyze how different
Santhoshi
The Concerned Parent
Situation
A viral but baseless article claims a new study found that a common vaccine causes severe side effects in children. The concerned parent, fearing for their child’s health, might choose to avoid vaccinations, increasing the risk to their child and others from diseases like measles.
News Article
Copying link to the article
Pastes in Debunked
Gets Verification with explanation
Social Media Channels
Himavanth
Whatsapp warrior
Situation
A misleading post shared in a WhatsApp group falsely states that a local political candidate plans to shut down several community centers, leading to a significant backlash from the community against the candidate based on false premises.
Article in WhatsApp
Social Media Channels
Forward it to Debunked WhatsApp
Open link in browser
Gets Message verification
Debunked Extension popup
Forward it to the source WhatsApp group
Gets Verification with explanation
News Article
Share via options
Akshaya
Company VC
Situation
Misinformation circulates on social media that the company is facing a lawsuit for a serious violation that never occurred. The VP may face pressures from stakeholders and reduced employee morale, and the company's market value might suffer until the misinformation is corrected.
Concerning post
Social Media Channels
Checks Debunked
Reports to Debunked
Article not found !
Hires Private Investigation
Forward it to the source WhatsApp group
Gets Verification with explanation
News Article
Share via options
breather Moment
before we get to screens
…
Misinformation awareness and debunking fake news service
Introducing
De
BunK
eD
Website
Landing Page
Santhoshi
The Concerned Parent
Santhoshi has been using debunked ever since he saw Debunked’s Insta channel post a reel about fake news related to child trafficking.
She uses Debunked to read up on news related to parenting, education and raising a kid. Her home page if full of verified news about the topics she desires.
It’s become a routine of his to play the fallacy game everyday, She learns about the various ways to spot misinformation.
She opens articles up and goes for the least sensationalized, most original news article. Thereby avoiding the hyper sensationalized news articles
Each article gives a brief about the topic, without sensationalizing any information.
Justifies the method of verifying the story
Fake News State
Verified News State
News analysis lists articles from other news sources, highlighting aspects like Trustability and Content originality of the articles
Which helps her know what to believe in
When she wants to learn more about risks of covid vaccine, she searches it up in the news analysis page to get a broad understanding of the various topics that have been talked about in this field
Web
Extension
She uses the Debunkd extension that automatically hides fake news and detects ai images and deepfakes.
Demonstration of Browser extension at work in articles
Himavanth
Whatsapp warrior
Himavanth came to know about Debunked when someone shared an article from the platform on the group. Since then he’s been reading the articles on the platform.
Himavanth came to know about Debunked when someone shared an article from the platform on the group. Since then he’s been reading the articles on the platform.
Whenever he feels like something fake has been sent on the WhatsApp group, he uploads the screenshot of it onto the platform.
Each article gives a brief about the topic, without sensationalizing any information.
Justifies the method of verifying the story
Each article gives a brief about the topic, without sensationalizing any information.
Justifies the method of verifying the story
The platform analyses the screenshot and verifies the news for him. Moreover, it explains how the news was debunked and what fallacy was used in the article to promote the news amongst people.
Himavanth uses the share as explanation feature to directly send a response to the fake news on the group. He asks the built in ai to change the tone of the explanation to make it sound non judgemental.
Nowadays since he’s discovered the Debunked business account on WhatsApp, he simply forwards his message to the business account and gets quick automatic debunking
Whatsapp Replies
Automated WhatsApp replies on the news article will be available
Himavanth will forward these to his community group to justify.
Direct link to the DBNKD site for more details.
Voice Assisted Debunking
Google assistance allows third party apps and service to work with
their voice assistant commands.
If Himavanth opens any article, he can ask google assistant
to "Debunk" the article.
With the voice assisted method, the justification is way easir to access.
Akshaya
Apollo VC
Akshaya works in a popular rubber processing company as a vice president, She has set up reminders on many news analysis topics that are crucial to her company’s industry (WIP).
How Its Debunk ?
News analysis lists articles from other news sources, highlighting aspects like Trustability and Content originality of the articles
Which helps her know what to believe in
Future scope and limitations
What's Next
Still got the sprit to!
A majority of news reading is done on mobile platforms, we need to simplify the features and create a mobile version.
The product still uses algorithms to suggest news articles, which may still lead to hyper partisanism. this need to be worked on further.
Explore govt subsidies and funding options for misinformation related enterprises
Illustrations By casualpolarbear
Check out these projects.
Product Design
Interaface
Desktop
Star Health
Web App
UI
Intern
Medlock+