Original Article

Ethical Concerns in the Use of Artificial Intelligence in Indian Online News Media

 

 

INTRODUCTION

The idea of artificial intelligence (AI) emerged in the 1950s and was successfully presented at the Dartmouth Conference. Later, research declined due to the lack of computer systems capable of processing the required amounts of data for the project to move forward. However, the idea was once again brought back during the 1990s and 2000s, when the level of technology had risen to a point where it could facilitate AI research. This time AI was made while mimicking the natural nervous system present within the human brain while simultaneously creating multiple systems like speech pattern recognition, content mimicry, visual recognition etc to finally combine to form a single system that can be called as an AI. These technologies can now create systems that generate text, recommendations, and automated content across numerous digital platforms.

Online news consumption in India had grown by a large margin compared to before with users relying on smartphones and social media platforms to acquire information in a daily basis. The information can range from daily news to other miscellaneous information.  The Reuters Institute Digital News Report 2025 indicates that a major proportion of Indian news viewers use AI chatbots such as ChatGPT and Google Gemini for news summaries and answers to questions ranging from simple information for fact checking to academic level information for help in their studies, this showed a widespread use of AI tools in daily information acquisition Chakrabarti et al. (2025). As AI collects, provides and controls editorial processes and audience’s access to information, ethical concerns about transparency, fairness, and accountability have become more and more apparent as each day passes.

This study examines these ethical concerns within the Indian online news, focusing on how AI-driven news systems had influenced visibility, trust, and public perception from 2020 to 2025.

 

Research Problem

The increased use of AI in Indian online news media has introduced ethical concerns related to algorithmic bias, misinformation or fake news, lack of transparency, and lack of accountability of liberal usage of certain functions of AI. As automated systems influence news production and news distribution, it is essential to understand how these changes affect journalistic integrity and public trust.

 

Research Gap

While global research has addressed the technological and ethical effects of AI in journalism, there is limited focus on the matter of Indian digital news ecosystem, particularly over the recent period marked by the rapid adoption of AI tools. Although studies discuss algorithmic bias and misinformation, there is a lack of India-focused analysis on how these ethical challenges evolve in the context of Indian online news platforms between 2020 and 2025. Though the study is being done currently for this paper, it is still based on the publicly available information from Indian online news media. The credibility of the news itself is brought under question as the chances of certain problems caused by AI to be hidden from public is very high

 

Research Questions

1)     How has artificial intelligence been integrated into Indian online news media between the years 2020 and 2025?

2)     What ethical concerns arise from AI-driven news systems, particularly in relation to algorithmic bias, misinformation/ fake news, lack of transparency, and accountability for certain unrestrained functions of AI?

3)     How do AI-driven tools influence news visibility and public opinion in India?

4)     What measures can be adopted to ensure ethical and responsible use of AI in digital journalism?

 

Objectives of the Study

·        To examine the use and impact of AI in news production and acquisition in Indian online media.

·        To identify ethical concerns associated with AI-driven news practices.

·        To evaluate how developments in AI systems affect journalistic integrity and public trust.

 

Hypothesis

The use of artificial intelligence in Indian online news media introduces ethical challenges, such as algorithmic bias, misinformation propagation, and lack of transparency, all of which are caused by the current AI technology’s lack of adaptation similar to that of human beings which may negatively affect journalistic credibility, scanning and processing of information and public trust without responsible ethical frameworks and oversight.

 

Review of Literature

The incorporation of artificial intelligence within journalism has given rise to a massive academic discussion with the major issues including automation, bias, and ethical responsibility in the use of AI, leaving the understanding of AI in the hidden matters of information, double meanings etc to be inadequate. AI’s power to read the hidden meanings has not reached the level of human beings which creates multiple problems when it comes to the matter of handling critical information that is used by both public as well as private sectors.

In “The Robotic Reporter: Automated Journalism and the Redefinition of Labor,” Carlson (2015) points out that automation changes the way newsrooms operate as it passes routine reporting tasks to algorithms which results in the redefinition of journalistic labor and authority, which is a necessary part of training that every journalist undergoes to understand and learn many hidden structures present within their job Carlson (2015).

Just like that Diakopoulos (2019) in “Automating the News: How Algorithms Are Rewriting the Media”, Goes into detail about the ability of algorithmic computer intelligence systems to not only functions as tools for productivity that aid the user in finding information but also be capable of shaping the way the receivers of the information perceive it through the codes that are built within them. Implying that the codes can in a way misinterpret information as sometimes they are unable to adapt with respect to the information Diakopoulos (2019).

The concern raised about AI systems which are not trustworthy and inherently biased because the codes and values embedded in them are not in line with certain communities and cultures is portrayed in “Reimagining Algorithmic Fairness in India and Beyond, ” where Sambasivan et al. (2021) shows us how information that is based on old values leads to an increase in societal inequalities and consequently a rise in the level of inequality and fights due to opposing cultures and values in society. Basically, pointing out that the codes in built within the current AI cannot keep up with the everchanging world where societal rules and values are changed everyday so as to reduce inequality. When AI follows rules of a certain value and the value changes next day, until the codes within the AI are changed, the AI will act upon this outdated values which in turn might have a change in increase of inequality.

His research demonstrates the necessity for the establishment of local AI rules that adhere to the cultural values and norms of that particular locality Sambasivan et al. (2021).

Reynolds and Nolan (2025), in “Ethical Considerations in AI Journalism: Bias Detection and Mitigation,” bring the discussion even further by stressing the significance of performing periodic algorithmic audits, being transparent, and using diverse datasets as measures to lessen and if possible, eliminate systemic bias in AI-generated news production. With the change of bias with respect to time, words and certain meaning would be rejected as time passes in society, regular audits will be helpful in making sure AI systems are up to date with the societal values Reynolds and Nolan (2025).

The ethical implications of AI-generated misinformation are widely discussed in contemporary research. In “AI Generated Fake Images and Videos in Misinformation Campaigns,” Nguyen (2020)analyze how generative technologies enable the creation of highly realistic synthetic media, complicating verification processes Nguyen (2020). Likewise, Shu et al. (2019)., in “Fake News Detection on Social Media: A Data Mining Perspective,” explore computational approaches to identifying misinformation while acknowledging the limitations of automated detection systems Shu et al. (2019). both are pointing towards the fact that the increase in the development of AI in creating of fake videos and information is becoming more and more foolproof as time passes which in turn will make it more difficult to prove truths using media evidences.

Further, Zhao and Phakdeephirot (2024), in “Exploring the Ethical Implications of AI-Driven News Production,” discuss the ethical dilemmas faced by journalists working with AI-integrated systems, particularly in relation to verification, accountability, and editorial control Zhao and Phakdeephirot (2024). Complementing this, Sonni et al. (2023) in “Artificial Intelligence in Journalism: A Systematic Literature Review,” synthesize existing research to highlight recurring concerns such as transparency deficits, data privacy risks, and the erosion of editorial accountability Sonni et al. (2023). using of Ai reduces the heavy workload taken up by journalists when it comes to matter of gathering information for any specific topic. But unlike human beings who can in a way detect certain false information just by reading it, AI systems have not yet become advanced enough to do the same. This in turn makes it difficult for journalists to trust the data gathered by AI as the publishing of one false information might destroy the life of an innocent person.

Recent scholarship also focuses on audience perception and algorithmic influence. Simon (2022), in “The Relevance of Algorithms: Public Perception and Automated News Selection,” examines how audiences perceive algorithmically curated news, revealing tensions between convenience and trust Simon (2022). Similarly, Gillespie (2014) work on algorithmic gatekeeping highlights how platforms and algorithms increasingly determine the visibility of information, thereby shaping public discourse and influencing democratic participation. Both of these people are pointing out to the fact that the convenience brought by AI when it comes to compiling and acquiring information for a topic from multiple sources at the same time makes more and more people depend on it to acquire information. But the inbuilt codes in the AI has the chance of filtering certain information before presenting it to the public, thereby creating a misinformed idea within the user which has a high chance of manipulating public’s political and democratic perception.

In the Indian context, emerging reports such as “Indians Biggest Consumers of AI-Generated News and Most Comfortable With It — Reuters Institute Report” indicate the rapid adoption of AI-driven news tools among Indian audiences, suggesting a shift toward automated information consumption Chakrabarti et al. (2025). Policy-oriented discussions, including “India Proposes Strict Rules to Label AI Content Citing Growing Risks,” reflect growing regulatory attention to misinformation and transparency in AI-mediated communication Kalra and Vengattil (2025). These reports indicate that the current era is moving towards a path where the already existing search engines are going to be abandoned as AI provides a better convenience in acquiring information.

Together, these studies demonstrate that while AI enhances efficiency and scalability in journalism, it simultaneously introduces complex ethical challenges related to bias, misinformation, transparency, and the restructuring of media power. However, there remains limited focused research on how these issues specifically manifest within the Indian digital news ecosystem, particularly during the period between 2020 and 2025, which this study seeks to address.

 

Methodology

This study employs a qualitative analytical approach using secondary data. The data was collected from publicly available national and international news reports published between the years 2020 and 2025, along with many AI research papers focussing on ethical problems from the same period. The sources were selected based on how much it had influenced the Indian public and the problems they faced due to AI. The collected materials were analysed to identify recurring patterns of certain key issues which includes algorithmic bias, misinformation and deepfakes, transparency concerns, and the impact of AI on public opinion and news consumption.

 

Analysis / Discussion

1)  Algorithmic Bias and News Personalization

Algorithmic bias arises when news recommendation algorithms prioritize content based on user engagement or certain input parameters rather than balanced editorial judgment, potentially creating echo chambers and reinforcing existing preferences. This bias can be shown in many instances in many different ways, the examples include-

·        The article “Facebook Let an Islamophobic Conspiracy Theory Flourish in India Despite Employees’ Warnings,” published in TIME magazine on 1 November 2021, reports that Facebook’s algorithm promoted “Love Jihad” conspiracy content in India despite internal warnings. Its engagement-driven system gave greater visibility to inflammatory posts, allowing such narratives to spread widely before corrective action was taken. This highlights algorithmic bias, where platform design prioritizes user engagement over ethical responsibility, thereby contributing to communal polarization.

The “Love Jihad” narrative refers to the claim that Muslim men deliberately target women from other communities to induce religious conversion through romantic relationships. The widespread circulation of this idea has strained interpersonal relationships and intensified communal tensions in India. Although Facebook employees reportedly raised concerns and attempted to limit the spread of such content, the platform’s algorithm continued to prioritize posts based on popularity and user interaction. As a result, trending and emotionally charged content was amplified regardless of its harmful social implications. This demonstrates a structural limitation of the algorithm, which lacks the capacity to adequately evaluate the ethical consequences of the content it promotes.

·        "Amnesty Report Highlights AI Risks of Algorithmic Decision-Making in India," published May 26, 2024, by Medianama. This article covers Amnesty International's April 30, 2024 report on AI risks in governance, focusing on welfare delivery problems from an Al Jazeera investigation.

In Telangana's Samagra Vedika platform, the algorithm wrongly denied subsidized food grains to 67-year-old widow Bismillah Bee from a below-poverty-line family. It used flawed data matching to tag her dead husband as a current car owner, ignoring census records. This is algorithmic bias because the system is trained on skewed data that systematically fails marginalized groups like poor widows, predictably erring against vulnerable people and amplifying inequality through opaque, automated welfare decision

·        "In India, an algorithm declares them dead; they have to prove they're alive," published January 25, 2024, by Al Jazeera. This news article reports on Haryana's Parivar Pehchan Patra (PPP) welfare database wrongly marking living elderly and widows as deceased, halting their pensions.

In 2020, Haryana launched PPP to verify welfare eligibility by linking family data on births, deaths, income, and assets. The algorithm declared 277 elderly citizens and 52,479 widows "dead" over three years due to mismatched death records from flawed government databases, cancelling pensions for living people like 102-year-old Dhuli Chand. This is algorithmic bias because the system uses incomplete data that systematically fails poor/rural households—prioritizing erroneous automated flags over real proof—predictably excluding vulnerable beneficiaries through opaque rules, automating poverty instead of aid.

All these are examples of algorithmic bias and how it has affected the country in different forms. The first incident had created rifts between people due to the struggles of divisions of caste and religion. The second and third show the injustices faced by small communities due to bias created by the AI parameters. But this can’t be blamed on AI, as the Ai has still not reached the level where it can independently make decisions when it comes to these parameters. They are still coded and decided by human minds that created or are using that specific AI.

2) Misinformation and Deepfakes

AI-generated misinformation and deepfakes pose significant threats to news accuracy and public trust. They are not only able to influence a community as a whole, but also have the ability to destroy the lives of a particular individual. The examples include

·        "Deepfakes Target Lok Sabha Election 2024 On Social Media," published May 28, 2024, by Resolver.com. This article analyzes how deepfakes and AI-generated content surged during India's 2024 Lok Sabha elections, spreading misinformation on platforms like WhatsApp and YouTube to manipulate voters and stoke divisions.

The incident details political actors deploying deepfake videos of celebrities like Ranveer Singh and Aamir Khan falsely endorsing parties, plus manipulated clips alleging EVM rigging and biased Election Commission enforcement. Algorithms amplified this synthetic content for high engagement, gaining millions of views before debunking; this shows algorithmic bias as recommendation systems—trained on polarized Indian data—prioritize sensational fakes over facts, systematically boosting divisive narratives that sway elections rather than neutral information.

·        "90% Indians exposed to fake endorsements in 2025," published November 13, 2025, by Business Standard. This article discusses McAfee's report on AI-generated deepfake celebrity endorsements fuelling scams, with Indians losing an average of Rs 34,500 per victim.

The incident details cybercriminals using deepfakes of Bollywood stars like Shah Rukh Khan (top target) and Alia Bhatt to promote fake skincare products (42%), giveaways (41%), and crypto/trading schemes (40%). Victims encounter these on social media, clicking phishing links or buying bogus items; 90% of Indians saw such fakes, with 60% involving influencers. This is algorithmic bias as platforms' recommendation systems, trained on engagement-heavy Indian content, prioritize viral deceptive videos over verified info, systematically spreading scams to millions rather than flagging fakes.

·        "Not just money, deepfakes robbing people of dignity," published November 24, 2025, by The Times of India. This article reports on non-financial deepfake harms, including revenge porn and harassment cases in India.

The incident details Rashmika Mandanna's October 2023 deepfake video (face swapped on a British-Indian influencer's body in revealing clothes), viewed millions of times on Instagram before deletion, sparking outrage and FIRs. Similar cases hit celebrities like Sachin Tendulkar (fake emergency message) and Katrina Kaif, plus non-celebs facing morphed porn for extortion. Algorithms amplified these via trending feeds, showing bias by favouring sensational "viral" content over safety checks, systematically harming women and eroding privacy/dignity.

Computer-generated fake pictures and videos were one of the most dreadful forms of crime before the coming of AI. Once, creating such pictures required heavy human skill, so the number of people conducting such crimes was of a manageable amount. But the advancements in AI and deepfake technology have made it so that even a simple person with a smartphone is able to mimic the skills of an expert editor. This has increased the amount of deepfakes in which people’s images are used to create degrading and humiliating videos. Just like before, AI is just a tool that listens to its user. The way it is used is decided by the user themselves. But then again, it also raises the ethical question of whether AI users should be given as much freedom as it has right now. Or should some rules be programmed so that such degrading deepfake creation can be prevented in the future?

3) Transparency and Accountability in AI-Mediated Journalism

These efforts reflect policy responses to the need for greater transparency in how AI influences news production, curation, and dissemination. We can say that the first two topics are all about the need for accountability of AI. In the current era of the internet and communication, AI in itself is as dangerous as weapons of mass destruction, as AI can be used to change public perception, create fake news, so as to spread panic and manipulate many people who are ill-informed. The examples of these accountabilities include.

·        "Indian Newspaper Day 2025: The Convergence of AI and Journalism," published January 28, 2025, by IndiaAI.gov.in. This government article examines AI's role in Indian newsrooms for tasks like summarization and misinformation combat, while stressing policy needs for transparency and bias checks.

It highlights ethical dilemmas from AI reliance, including deepfakes worsening misinformation, urging stricter governance policies. This reflects transparency efforts as news organizations must disclose AI use in production/curation (e.g., NLP transcription, video summaries) to maintain trust; without mandated labelling and audits, opaque AI influences dissemination, risking unaccountable bias in public info—prompting calls for regulatory frameworks to enforce accountability in AI-mediated journalism.

·        "AI-generated journalism: Do the transparency provisions in the AI Act give news publishers enough protection?" published October 22, 2024, by Internet Policy Review. This analysis evaluates the EU AI Act's transparency rules for AI in news production and their implications for publishers.

The article examines how newsrooms use generative AI for automated summaries, translations, and personalization without disclosing it to readers. It shows a lack of accountability as opaque AI decisions shape content curation—what stories trend or get recommended—risking biased dissemination without human oversight labels. This reflects policy responses like mandatory AI watermarks and audit trails to enforce transparency in AI-mediated journalism, ensuring public trust amid growing automation.

These are just a few instances where accountability for the growing AI community was taken. As AI improves more and more, more and more questions are asked and more and more measures should be taken.

4) Influence on Public Opinion and News Consumption

AI systems and algorithmic personalization significantly influence news consumption behaviors and public perceptions. AI helps in the gathering and arranging of information from multiple sites, thereby decreasing hours of work to a few minutes. This in turn, makes people heavily rely on it for procuring information, thereby giving AI more power over the information obtained by each individual, and this in turn, gives AI influence on public opinion and news consumption. A few examples include.

·        "How audiences think about news personalisation in the AI era," published June 17, 2025, by Reuters Institute Digital News Report. This chapter analyzes how AI-driven personalization shapes news selection, formats, and engagement across 48 markets.

AI recommendation systems create filter bubbles by tailoring homepages, alerts, and summaries to past behaviour, boosting efficiency but reducing news diversity. Users like young audiences favor AI for relevant summaries/translations (top interest at ~30%), yet many distrust opaque algorithms seen as "less biased than editors" by some but engagement-driven by others; this influences opinion by prioritizing sensational content, fragmenting perceptions as platforms like BBC experiment with AI audio/text conversion, weakening direct publisher loyalty.

·        "The impact of AI in news delivery and how it shapes public opinion," published June 13, 2023, by IndiaAI.gov.in. This piece explores how AI personalization and NLP tools transform news access in India, influencing consumption patterns and perceptions.

AI algorithms analyse user behaviour to deliver tailored feeds, summaries, and regional translations, boosting engagement but risking echo chambers. Personalization favours familiar views over diverse ones, while NLP translation reduces language barriers yet may embed biases from training data; this shapes opinion by prioritizing engaging (often polarized) content, fragmenting national discourse as users in multilingual India get siloed narratives rather than balanced info.

·        "Prediction 2022: The year AI becomes human-focused," published March 31, 2022, by IndiaAI.gov.in. This article predicts AI's growing role in personalizing news delivery to match user preferences, influencing consumption and opinions in India.

AI analyzes clicks, search history, and location to curate feeds and summaries, making news more relevant but creating filter bubbles. Users receive reinforcing content over diverse perspectives, while NLP translations for regional languages may carry training biases; this shapes public opinion by prioritizing engaging stories, potentially polarizing views in multilingual India rather than fostering balanced discourse.

These are just public news reports on how AI has gained a major influence in the current internet-filled society. Its usage has risen so much that it rivals the usage of top search engines like Google. The influence AI has on the younger generation is even more prominent as most of the day to day works are easier with the assistance of AI.

 

Conclusion

The integration of AI into Indian online news media has enabled efficiencies in content production and personalization but has also introduced significant ethical concerns. Algorithmic bias, misinformation, and deepfakes, lack of transparency, and transformative effects on public opinion highlight the need for ethical frameworks, policy interventions, and editorial oversight. Developments between 2020 and 2025 reveal that while AI technologies have permeated news ecosystems, their responsible and ethical deployment requires transparency, accountability, and safeguards that protect credibility and uphold democratic discourse.

 

ACKNOWLEDGMENTS

None.

 

REFERENCES

Booth, R. (2025, November 5). The Chilling Effect: How Fear of "nudify" Apps and AI Deepfakes is Keeping Indian Women off the Internet. The Guardian.

Carlson, M. (2015). The Robotic Reporter: Automated Journalism and the Redefinition of Labor. Digital Journalism, 3(3), 416–431. https://doi.org/10.1080/21670811.2014.976412

Chakrabarti, A., et al. (2025, June 18). Indians Biggest Consumers of AI-Generated News and Most Comfortable with it—Reuters Institute Report. ThePrint.

Chatterjee, S. (2025). Experts Hail Deepfake Labelling Rules but Flag Overreach Risks. The Economic Times.

Diakopoulos, N. (2019). Automating the News: How Algorithms are Rewriting the Media. Harvard University Press. https://doi.org/10.4159/9780674239302

Gillespie, T. (2014). The Relevance of Algorithms. In T. Gillespie, P. J. Boczkowski, and K. A. Foot (Eds.), Media Technologies: Essays on Communication, Materiality, and Society (167–194). MIT Press. https://doi.org/10.7551/mitpress/9780262525374.001.0001

Kalra, M., and Vengattil, A. (2025, October 22). India Proposes Strict Rules to Label AI Content Citing Growing Risks. Reuters.

Nguyen, D., et al. (2020). AI-Generated Fake Images and Videos in Misinformation Campaigns. Journal of Digital Media.

Reynolds, S., and Nolan, J. (2025). Ethical Considerations in AI Journalism: Bias Detection and Mitigation. ITSI Transactions on Electrical and Electronics Engineering, 12(1). https://doi.org/10.65521/itsi-teee.v12i1.144

Sambasivan, N., Arnesen, E., Hutchinson, B., Doshi, T., and Prabhakaran, V. (2021). Reimagining Algorithmic Fairness in India and Beyond. In Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency 315–328). Association for Computing Machinery. https://doi.org/10.1145/3442188.3445896

Shu, K., Sliva, A., Wang, S., Tang, J., and Liu, H. (2019). Fake News Detection on Social Media: A Data Mining Perspective. ACM SIGKDD Explorations Newsletter, 19(1), 22–36.

Simon, F. M. (2022). The Relevance of Algorithms: Public Perception and Automated News Selection. Reuters Institute for the Study of Journalism.

Sonni, et al. (2023). Artificial Intelligence in journalism: A Systematic Literature Review. Journalism Studies.

Zhao, J., and Phakdeephirot, N. (2024). Exploring the Ethical Implications of AI-Driven News Production. Academic Journal of Science and Technology, 12(1), 69–75. https://doi.org/10.54097/8ewxk898