News, views, and AI: How bots threaten the future of news publishing

Raghav Bahl and Ritu Kapur’s The Quint began reducing its dependence on search in 2023. By the middle of 2024, it simply gave up on pushing its stories on Google through keywords.  

“We moved away from high-volume publishing, including wire copy, quick-turn videos, sports, entertainment, and lifestyle, to a more deliberate focus on opinion, analysis, ground reports, and investigations,” says Kapur, cofounder and managing director. “We are now a more niche, deeper offering, journalistically.” 

This has pushed up direct traffic  to about 28 per cent instead of the usual low single digits in general  news publishing. 

 

Whether it is The Hindu or Malayala Manorama, almost every strong editorial brand in India has chosen the direct, paid route to reach and monetise readers. They are now thankful for it, albeit for another reason. Readers coming through search, which traditionally brings in a bulk of the traffic for news, are dropping at an alarming rate as artificial intelligence (AI) summaries rise.   

 

From June 2024 to June 2025, the top 10 English language publications in India lost 20-30 per cent of their online audience, according to media analytics firm Comscore. Overall, people seeking news or information online dropped by about 5 per cent 

in 2024 over 2023, while the time  spent on news dropped by a third  (see charts). 

 

“The drop in ad revenue is likely to be worse, since higher reach commands a premium,” says Siddharth Jhawar, country manager, Moloco India, which provides machine-learning based advertising advice. 

Jing Shyang Yeow, director, client insights at Comscore Inc, says in the last three years, they have seen a drop in unique visitors to news and information sites globally, including to aggregator platforms such as Google News. Ask Kate Beddoe, managing director, news partnerships, Asia-Pacific at Google, and she insists, “Overall, total organic click volume from Google Search to websites has been relatively stable year-over-year.”

 

Whether it is Google offering AI summaries or OpenAI’s ChatGPT or Perplexity giving information gleaned from scraping publisher sites, the rise of AI has now become the biggest challenge to publishing since the advent of the internet. So, the focus on direct traffic and a deeper connection with readers is first among a combination of steps publishers across the world are taking to protect their audience and business.  

The use of software tools like Cloudflare or TollBit to either block AI crawlers/bots that scrape data from news websites or make them pay to do so is the second step.  

The third is to either talk to both Google and AI firms on fair compensation or take them to court. 

The compensation question 

Of the ₹70,000 crore spent on digital advertising in India in 2024, Google, Meta, and Amazon Retail Media bagged ₹50,000 crore, or over 70 per cent. The remaining ₹20,000 crore went to everyone else: Over-the-top (OTT) platforms, news publishers, local sites, information sources et al. Within them, bigger players such as JioHotstar and Times Internet pocketed the chunk, leaving hardly anything for smaller publishers.  

For Quint Digital, which earned  ₹65 crore in revenue in the 2024-25 financial year (FY25), it made no sense to keep doing stories on celebrity birthdays or the latest diet fad in the hope of appearing at the top of the search page. It chose to pivot away from search. 

Indian language publications, meanwhile, are relatively protected for the time being. “We are not seeing it (the impact of AI) yet,” says Mariam Mammen Mathew, CEO of Manorama Online, the digital arm of Malayala Manorama. That’s because AI hasn’t quite gotten around to Indian language publications so far. When it does, “no doubt it will hit us badly, so we are trying to concentrate on direct traffic,” says Mathew.  

Girish Agarwal, non-executive director, DB Corp, which counts Dainik Bhaskar among its publications, likens it to the time when Google arrived on the scene over 20 years ago. “All of us (publishers) put our content on Google. It became the holder of individual brands and their traffic,” he says.  

There was no compensation for the traffic news publishers attracted for Google, and ad revenues could  hardly make up for the readership shifting online.

 

Today, for most large newspapers, the print business, which continues  to be profitable, subsidises their online presence. “Now, AI has taken the place of search engines,” says Agarwal.  That means a further pressure on  print revenues. 

 

Beyond the worry about shrinking traffic and revenue is a bigger concern: the appropriation of content.  

“Byte-sized, snackable content usage is on the rise,” explains Kapur. “If I do something on, say, the issue with voter data and someone is searching for that, my article will be served as a summary, even if there is a link available,” she says. “The attention-deficit audience is not going to come to my article. So, the AI firm ought to pay me for that summary.”  

Wooing and suing 

The Digital News Publishers Association (DNPA), which represents 22 mainstream publishers (including this newspaper), approached OpenAI about six months ago. It proved to be a fruitless exercise, say several members who were part of that Zoom call with Varun Shetty, the New York-based head of media partnerships at OpenAI.  

“When we talked about attribution and compensation. They said, ‘We are a startup and can’t pay’,” says a person who was present at that meeting.  

Coming from a startup that is on the verge of being valued at $500 billion through its latest share sale, that was a bit rich, says one publisher. “You (OpenAI) are selling a subscription service from an asset created with our data,” adds another. 

ChatGPT is available at a monthly subscription of $20 (about ₹1,760) in the US. In India, it launched a ₹399 per month subscription in August. 

While a deal with Indian publishers is far from sight, OpenAI is engaged in hectic parleys in other markets.  

It will be paying an estimated $250 million to the US-based NewsCorp over five years for the use and display of content from a clutch of publications, including The Wall Street Journal and The Times. There is also a deal with Germany’s Axel Springer, owner of media outlets including Politico and Business Insider.  An e-mail to Shetty remained unanswered.

 

Google, on the other hand, is politely engaging with publishers through workshops and events, but has given little ground. “We believe that our use of publicly available information for training AI models is based on the legal principle of fair use,” says Beddoe. 

This is where the legal conflicts come into play. Late last year, ANI filed a lawsuit against OpenAI in the Delhi High Court. It alleged that OpenAI used ANI’s content to train its AI models and generated false information attributed to the news agency. DNPA and Indian Music Industry (IMI), the body representing music labels, soon joined 

the lawsuit.  

Across the world, publishers such as The New York Times, authors like George RR Martin and John Grisham, besides record labels and musicians are among those suing AI firms over copyright infringement. Their contention is that these firms are violating copyright laws by using existing, original or secondary published data to train the AI.  

In the US, in two cases, judges have ruled in favour of AI firms. Meanwhile, a report on copyright and AI in May this year by the US Copyright Office noted: “Making commercial use of vast troves of copyrighted works to produce expressive content that competes with them in existing markets, especially where this is accomplished through illegal access, goes beyond established fair use boundaries.” Days after the report was published, American  President Donald Trump fired Shira Perlmutter, the head of the US  Copyright Office. 

 

In India, publishers like DB Corp and Malayala Manorama have started identifying and stopping AI bots from crawling their data. But it’s not an easy task.

 

“You block 100,000 AI crawlers, and the next day there are more,” says Sujata Gupta, secretary general, DNPA. “Even the algorithmic change in the crawlers is automated.” Referral traffic still ensured some credibility in the source, but if it is ChatGPT, then the source can be anything, she adds. “You are harming the discoverability of  good journalism,” Beddoe counters this argument: “The demise of journalism has been predicted with every new technological wave, but it manages to survive and find new audiences and ways of connecting.” 

 

Yeow reckons that news publishers will continue to be challenged if their content is purely written/text. Not surprising then that podcasts and videos from news publishers have multiplied over the last few years.  

Comscore’s social incremental data shows that “publishers have partially offset losses in direct traffic to their sites and apps with significant presence and reach on platforms including Facebook, Instagram and X in the India market,” says Yeow.

That said, what the AI challenge will do to revenues is a question no one has the answer to at this point.