Are AI-generated summaries the wake-up call for digital publishers to evolve?

New Delhi: The way we consume content is changing rapidly, especially now with the functionalities and integration of GenAI expanding across sectors. 

Tech giants like Google (Bard summaries within Google Search), Meta (Facebook AI's research on summarisation tools), and Microsoft (Project Bonsai for code summarisation) are developing features that automatically generate summaries of information from websites. These tools cater to busy audiences seeking quick overviews. 

With users obtaining answers directly from the search results using GenAI, there is a fear among publishers that users may not avail themselves of the clickable links to visit their websites, resulting in decreased traffic. 
In May 2024, Google launched 'Overview' in the US, offering users summaries without requiring website navigation.

However, it swiftly rolled back the feature amidst criticism over inaccuracies. As Google continues to work on this feature, users are meanwhile making use of MetaAI for its integrated function of providing summaries to user prompts. 

Like two sides of a coin, these AI-led changes will have both pros and cons. 

For digital publishers relying on ad revenue from in-depth articles, AI summaries pose a threat. Readers might bypass detailed content for the AI-generated gist, leading to a decline in readership and ad revenue.

Deepak Nagpal, Executive Editor of Outlook India (Digital), shed light on the topic and said, “With AI growing in significance in digital media and providing such overviews and snippets, the impact on digital publishing will be immense.” 

On the other hand, imagine a farmer in a remote village struggling to understand a government agricultural subsidy programme website. An AI summary, translated into their local language, could be a game-changer for information access. 

Pradeep Gairola, Head of The Hindu Group (Digital), opined, “The sheer number of specialist AI properties is increasing at an astronomical pace.” 

He compared the evolving AI market to a shift from a 'one-stop-shop' approach to a more specialised model. Consumers are increasingly seeking niche content, and AI tools are likely to follow suit.

He highlighted the evolution of social media platforms as islands as the cause of the shifting focus of platforms like Facebook, which prioritises news content below entertainment formats like that of celebrity news stories.

Nagpal points out that this might push reporters to focus on more quality content, including strengthening ground reporting and fact-checking, as processing real-time and up-to-date data is still a challenge for Gen-AI chatbots. 

Another concern Nagpal highlights is revenue-sharing models, as GPT tools continue to train on copyrighted content. While industry bodies might be exploring ways to capitalise on leveraging their content, he opined that “by doing videos and earning revenue from YouTube, from Instagram, Facebook, or whatever other platform-oriented ways, the revenue gets shared within that platform.” 

A similar approach can become a necessity in the near future if GenAI-led transformations get judiciously integrated with the operational processes of news publishers.

As disruptions and innovations give way to evolution, adaptability is important. This is why even digital news publishers might have to take a look at how AI-led changes can complement their efforts. 

“They could make use of technology for better targeting of the audience, especially when it is about catering to a niche base. Even in cases otherwise, the current AI-led disruption will push news publishers to go the extra mile to take cognizance of how they can build specialised offerings (considering news as a ‘commodity’ in this case),” said Nagpal.

Noting the diversity that news platforms cater to, Garoila believes that it’s still too early to fully grasp the overarching impact of AI on the industry.

He added, “AI is operating on a perplexity model, where after a prompt is sent to them, they give the answer and give you the sources from which that answer has been made.”

He stressed on the other side of the story that after the ‘prompt-response’ cycle is over, publishers have to really figure out if generic news is a commodity and question the value addition it gives to the consumer. 

He also added, “We need to look at the change in content consumption because every time Google or Meta rolls out a new feature, they are not doing away with the old set of features immediately; they are also observing. Meanwhile, we have to focus on quality content, personalisation, and the correct targeting of the audience.” 

There is a likelihood of AI tightening its grip on written content, but its limitations in understanding regional languages create an opportunity for short-form video content. Unlike current AI models that struggle with regional languages, video content transcends these limitations. This format leverages the power of visual storytelling and resonates with a mobile-first audience. Additionally, short videos are easily shared across social media platforms, maximising reach.

Vishesh Sharma, FinTech Marketing Leader, said, “This is like a boost to the sea of choices we can make when it comes to adapting. Again, it's democratising the entire process of serving content to your consumers.” 

He even observed that with quality video-form content, there are chances for the algorithms to push the content further in terms of reach and visibility.

Here, AI summaries and short-form video content can co-exist and even complement each other. Imagine a financial website offering AI summaries of complex investment plans in Hindi, accompanied by short explainer videos with visuals and audio. This caters to both quick-glance seekers and those wanting a deeper dive.

However, Sharma also emphasised the glass-half-empty side of things as he stressed the importance of the right kind of automation due to the diversity of languages that India caters to. 

AI features, therefore, could be a great push for digital publishers to focus more on a strong short-form video content strategy, creating a seamless user experience. Simultaneously, this allows users to switch between formats easily across platforms.

Amid the sea of AI-led transformations, small-scale publishers will have a difficult time adjusting to the disruptors. 
“Small-scale businesses neither have the kind of resources nor the kind of might, energy, or effort to disrupt their business and still make money. The room for any disruption is very limited for them,” said Sharma.

Nagpal highlighted that while GenAI integrations are increasing fear of user privacy, disinformation, misinformation, and echo chambers, there are also risks of small-scale publishers or bad actors relying on AI for quick results. This would either mean publishers peddling information's authenticity or ignoring fact-checks.

Looking forward

Nagpal expressed that AI is going to become part of our daily lives. While one cannot be dependent on it for the collection of facts or sourcing of information, Nagpal emphasised that industry bodies have to come together and connect with these AI tool developers to find a middle path where both can co-exist in harmony.

Garoila emphasised the inevitability of change, noting that while general news is often viewed as low-value, personalised and in-depth news that connects dots and offers interpretations holds greater value for users. Depending on their size and ambitions, news organisations must strategically choose their focus. Niche outlets should shift from low-value, commodity news to high-value, personalised content, while larger organisations must navigate a balance: general news attracts a broad audience for ad revenue, whereas personalised news drives subscriptions.

The key takeaway is that while AI summaries might disrupt written content, Indian digital publishers can adapt by focusing on high-quality, in-depth content with strong fact-checking and insightful analysis. This will ensure content remains valuable and trustworthy, even in the age of AI-generated summaries.