{"id":9895,"date":"2026-01-22T08:00:00","date_gmt":"2026-01-22T13:00:00","guid":{"rendered":"https:\/\/cnti.org\/cnti-news\/\/"},"modified":"2026-03-25T16:01:16","modified_gmt":"2026-03-25T20:01:16","slug":"case-studies","status":"publish","type":"post","link":"https:\/\/cnti.org\/reports\/chatbots-for-news\/case-studies\/","title":{"rendered":"Case Studies"},"content":{"rendered":"\n
<\/div>\n\n\n\n

Case Study 1: \ud83c\uddee\ud83c\uddf3 The skeptical power user<\/strong><\/h2>\n\n\n\n

This interviewee represents the archetype of the \u2018skeptical power user,\u2019 one who distrusts traditional television news due to perceived bias but embraces the conversational power of AI chatbots<\/span><\/span>. They turn to AI chatbots for all four kinds of information needs we have covered in this report: the need to act, the need to know, the need to understand and the need to feel.<\/p>\n\n\n\n

News repertoire<\/span><\/span> and sources: Synthesizing ideological balance<\/strong><\/h4>\n\n\n\n

This interviewee\u2019s news consumption is characterized by a rigorous pursuit of ideological balance and a rejection of sensationalism. They avoid Indian television news entirely, citing bias, and instead prioritize international news outlets like Al Jazeera<\/em> and domestic English language print sources such as The Hindu <\/em>and The Times of India<\/em>. In their hierarchy of trust, written ground reports rank highest, followed by AI chatbots. While they access information via social media and YouTube, they rank YouTube videos last in terms of reliability, viewing them as prone to biased propaganda. When researching complex topics like the conflict in Gaza, they typically consume 10\u201312 articles and one to two videos from “both sides” to synthesize a neutral view, and hold “whole conversations with ChatGPT,” rather than relying on a single narrative.<\/p>\n\n\n\n

AI chatbot usage and information needs: Controlled delegation<\/strong><\/h4>\n\n\n\n

They delegate a wide array of tasks to AI chatbots, ranging from tracking daily macronutrients to debating sports statistics, yet retain ultimate authority by verifying the AI chatbot\u2019s outputs against traditional news sources. Their relationship with AI <\/span><\/span>technology is highly personalized. They nicknamed their preferred AI chatbot, ChatGPT, “Curler” (referencing FC Barcelona fans) and self-deprecatingly refer to themselves as “Burden” due to their relentless questioning. They instruct Curler to be “blunt” and “unbiased.”<\/p>\n\n\n\n

AI chatbot preferences: \u201cA conversation between two people\u201d<\/strong><\/h4>\n\n\n\n

They display a strong preference for ChatGPT over alternatives like Gemini, stemming largely from the former\u2019s conversational fluidity. They feel ChatGPT’s answers are “easy” and feel “like a conversation between two people,” whereas they described Gemini\u2019s outputs as “way too hard to respond” <\/em>and “like AI AI.” <\/em>Their comfort with ChatGPT extends to data sharing (e.g., name, birth year, family history), whereas they use Gemini only sporadically. <\/p>\n\n\n\n

Trust and verification: Trust, but verify <\/strong><\/h3>\n\n\n\n

While they find AI chatbots to be “mostly right,” they do not exhibit blind trust. Their workflow involves an iterative verification loop: searching a topic on ChatGPT, clicking linked sources to read full articles, then comparing the text against the AI chatbot\u2019s summary to ensure accuracy. They retain ultimate authority, using the AI chatbot to distill information but relying on established news brands for ground truth.<\/p>\n\n\n\n

They also acknowledge that the AI chatbot is sometimes wrong, occasionally failing to provide all the information found in the full linked article. When they notice the AI chatbot making a mistake, they correct it and instruct it to \u201cdouble check\u201d its data and learn from the error for future reference. <\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n
<\/div>\n\n\n\n

This interviewee was interested in knowing more about the latest updates in the ongoing War in Gaza. They started off with a broad query and followed up with more specific questions on the hostage situation and Hamas\u2019 statement. They found the links provided with every response very useful. Typically, they\u2019d click on one of the links from established media outlets like Al Jazeera<\/em>, The Times of India<\/em>, the Wall Street Journal<\/em> and CBS News<\/em> and go through the articles. They employed a structured information pipeline: search, click on the AI chatbot sources, read the articles in their entirety and use the information to ask follow-up questions.<\/p>\n\n\n\n

<\/div>\n\n\n\n
\n\n\n\n
<\/div>\n\n\n\n

Case Study 2: \ud83c\uddee\ud83c\uddf3 The strategic snacker <\/strong><\/strong><\/h2>\n\n\n\n

This interviewee represents a strategic \u2018snacker\u2019 of information. They exhibit a hybrid news consumption style that blends traditional print media with digital news sources and use AI chatbots to filter noise and curate a personalized news feed. <\/p>\n\n\n\n

News repertoire and sources: Editorials, op-eds and summaries<\/strong><\/h4>\n\n\n\n

Their news repertoire is anchored by subscriptions to physical newspapers, specifically English language ones like The Hindu<\/em> and The Indian Express<\/em> and the Punjabi version of The Tribune<\/em>. They value the tactile experience of reading physical papers, particularly the editorial and op-ed sections, which they use to improve their vocabulary and gain deeper knowledge. For scanning headlines and quick, daily updates, they rely heavily on Inshorts<\/em>, a popular Indian news app that summarizes news stories from various sources into short, sixty-word bites. They also access news via social media and occasionally visit newspaper websites when away from home.<\/p>\n\n\n\n

AI chatbot usage and information needs: A \u2018wide net\u2019 prompting strategy<\/strong><\/h4>\n\n\n\n

To meet their information needs, the interviewee integrates AI chatbots like ChatGPT and Gemini into their routine, using them almost every day. Their primary motivations for using AI chatbots are time efficiency and the ability to summarize vast amounts of information quickly. They employ a \u2018wide net\u2019 prompting strategy, typically asking the AI chatbot to generate lists (see example). From these generated lists, they select specific stories of interest to dig deeper through follow-up requests for detailed summaries. They finalize this process by asking the AI chatbots to condense information such as dates and headlines into a concise paragraph to explicitly aid in memory retention.<\/p>\n\n\n\n

AI chatbot preferences: ChatGPT, the Go-to App<\/strong><\/h4>\n\n\n\n

They primarily use ChatGPT and Gemini on mobile, while occasionally using Claude and Perplexity on their laptop. They consider ChatGPT their go-to app, noting they have used it the longest, find it easy to use and subscribe to the paid version of the app. Notably, they prefer Gemini for its \u2018deep research\u2019 feature, which they utilize for academic projects and reports. They specifically value Gemini’s ability to produce organized outcomes, such as proper PDFs and flowcharts. In direct comparisons, they noted that Gemini often provides a better information hierarchy with useful sub-points, whereas ChatGPT sometimes offers better general overviews for broad topics.<\/p>\n\n\n\n

Trust and verification: Diverse curation<\/strong><\/h4>\n\n\n\n

The interviewee generally trusts AI chatbots to provide a balanced view by curating information from multiple sources, contrasting this with the singular perspective of a specific journalist. However, they remain skeptical in specific contexts. For instance, they expressed distrust toward Indian government data regarding politically contentious topics like climate change, preferring international sources instead. Furthermore, within their professional domain of law, they have detected inaccuracies where the AI chatbot fabricated facts about case laws, which led them to verify such data against books and Google searches.<\/p>\n\n\n\n

This example shows the interviewee engaging in “snackable” consumption of news, prioritizing breadth over depth. Their interaction patterns are characterized by frequent requests for “Top 10 international news,” “Top 5 sports news”<\/em> and similar short lists.<\/p>\n\n\n\n

\n
\n
\n
\n

Indian user<\/p>\n\n\n\n

\n

Today news<\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n

\n
\n
\n
\"\"<\/figure>\n\n\n\n

ChatGPT<\/p>\n<\/div>\n\n\n\n

\n

Today\u2019s major headlines<\/strong><\/p>\n\n\n\n

1. India-US Trade Deal Nearing<\/strong><\/p>\n\n\n\n

India and United States are reportedly close to a trade agreement that would reduce U.S. tariffs on Indian imports to about 15-16% (down from much higher levels. \ud83d\udd17<\/p>\n\n\n\n

Why this matters:<\/strong><\/p>\n\n\n\n

\u2022 Lower tariffs can boost Indian exports to the US, helping Indian industries.<\/p>\n\n\n\n

\u2022 For the U.S., It signals a potential shift in trade strategy with India.<\/p>\n\n\n\n

[…]<\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n

\n
\n

Indian user<\/p>\n\n\n\n

\n

Give me today\u2019s major in international news<\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n

\n
\n
\n
\"\"<\/figure>\n\n\n\n

ChatGPT<\/p>\n<\/div>\n\n\n\n

\n

Here’s one major piece of international news today:<\/p>\n\n\n\n

Kyiv, Ukraine: Russian Missiles & Drones Strike<\/strong>
\u2022 Overnight, Russian missile and drone attacks struck Kyiv and surrounding regions, killing six people<\/strong> (including two children<\/strong>) and causing widespread power outages across Ukraine. \ud83d\udd17<\/p>\n\n\n\n

[…]<\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n

\n
\n

Indian user<\/p>\n\n\n\n

\n

Can you give me a detailed summary about the Kyiv Ukraine conflict<\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n

\n
\n
\n
\"\"<\/figure>\n\n\n\n

ChatGPT<\/p>\n<\/div>\n\n\n\n

\n

Absolutely – here’s a detailed, clear summary<\/strong> of the Kyiv-Ukraine conflict (Russia-Ukraine war)<\/strong> as of October 2025<\/strong>, covering its origins, timeline, key developments, and current status.<\/strong><\/p>\n\n\n\n

[…]<\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n

\n
\n

Indian user<\/p>\n\n\n\n

\n

Give me a paragraph with all important dates and headlines<\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n

\n
\n
\n
\"\"<\/figure>\n\n\n\n

ChatGPT<\/p>\n<\/div>\n\n\n\n

\n

Here’s a compact paragraph<\/strong> summarizing the Russia-Ukraine conflict with all the major dates and key headlines:<\/p>\n\n\n\n

The Russia-Ukraine conflict began in 2014<\/strong>, when Russia annexed Crimea<\/strong> and separatists seized parts of Donetsk and Luhansk<\/strong>…<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n

<\/div>\n<\/div>\n\n\n\n

When asked how they feel about the responses, the interviewee said,<\/p>\n\n\n\n

\n

I don’t think there are more than ten important news stories daily on a national level, so I think usually one or two or three are most important. So if there’s something that I have a little bit of interest in, then I’ll tell them that. ‘Give me a detailed summary of it,’ or ‘Give me a detailed version of it.’<\/p>\n<\/blockquote>\n\n\n\n

<\/div>\n\n\n\n
\n\n\n\n
<\/div>\n\n\n\n

Case Study 3: \ud83c\uddfa\ud83c\uddf2 The perspective-seeker<\/strong><\/h2>\n\n\n\n

This interviewee relies more heavily on AI chatbots for news than anyone else CNTI interviewed in the U.S. They regularly go to Gemini for a quick roundup of headlines, although they also value \u201cquality time with news\u201d <\/em>when they need to learn about something with direct personal impact, like understanding the pros and cons of a local ballot proposition.<\/p>\n\n\n\n

News repertoire and sources: Seeking out multiple ideological perspectives<\/strong><\/h3>\n\n\n\n

This interviewee uses a broad repertoire of sources, including news outlet apps with regular push notifications, the social media feeds of friends and family members, conversations with coworkers and Google Gemini. They said they are concerned about bias, but are also particularly interested in getting news from multiple ideological perspectives. Following people they knew on social media means they can evaluate the slant based on what they knew about that other person. They also find it particularly important to get news that informs their individual actions. Specifically, they turn regularly to Gemini for round-ups of what was happening in their zip code and what topics will be on their ballot. <\/p>\n\n\n\n

\n
\n
\n
\n

U.S. user<\/p>\n\n\n\n

\n

What’s the latest headlines in [zip code]?<\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n

\n
\n
\n
\"\"<\/figure>\n\n\n\n

Gemini<\/p>\n<\/div>\n\n\n\n

\n

The ZIP code [zip] covers the central and downtown area of [city].<\/p>\n\n\n\n

Here are some of the latest headlines and recent local news:<\/p>\n\n\n\n

Recent Crime & Public Safety:<\/strong><\/p>\n\n\n\n