AI is everywhere. Amidst all the doom and hype, newsrooms are discovering a wide range of use cases for AI in journalism and its workflows. However, as with any emerging technology, this often leads to the mindless application of tools.

For instance, in India, TV news channels are now filled with AI anchors. In a market notorious for misinformation, bigotry, and government propaganda, introducing an AI anchor does little to address the trust deficit among viewers. This is what I call “using AI for AI’s sake”— employing a tool not to solve a problem or benefit users, but merely as a flashy novelty.

That said, globally, there have been excellent examples of AI applications in newsrooms — streamlining workflows, automating tedious tasks, and aiding investigative journalism. AI is already becoming an integral part of newsroom operations and product development.

However, one area where AI is underutilized — and one I believe will gain traction in 2025 — is in making news products more accessible.

Thanks to advances in assistive technology and the efforts made by tech giants to bake accessibility into their design, much of the internet is, by and large, accessible. Yet most news platforms continue to be inaccessible. News platforms in the Global South, in particular, often fall short. Even in the U.S. and Europe, accessibility is frequently treated as a bare-minimum compliance measure to avoid lawsuits, rather than as an essential feature.

There are several reasons accessibility remains sidelined in newsrooms, regardless of size, region, or language. These include weak regulations, lack of awareness among leadership, and (especially) resource constraints. Many newsrooms, large and small, lack the bandwidth for both the technical overhaul required to ensure accessibility and the day-to-day tasks that support it, such as adding captions, transcriptions, and alt text.

These challenges aren’t unique to smaller newsrooms. Even some of the largest outlets in the Global South struggle with accessibility — not for a lack of intent, but because retrofitting outdated tech stacks to meet accessibility standards can be a monumental task. On the other hand, newer newsrooms may prioritize basic accessibility features but lack resources for subtitling podcasts or crafting high-quality alt text. This is where AI can step in to bridge the gap.

Earlier this year, we overhauled the Newslaundry website to make it accessible. Beyond technical upgrades, we integrated AI into our production and publishing workflows, automating tasks like generating subtitles and transcriptions for podcasts.

During my time as part of the inaugural AI Journalism Lab at the Craig Newmark Graduate School of Journalism, I was exposed to AI thought leaders and cutting-edge experiments in the field. This inspired me to explore how AI could be leveraged to make news products more accessible. Our accessibility efforts didn’t just make our platform more inclusive — they also attracted new paying subscribers. This demonstrated that AI-driven accessibility isn’t only better product design but also good business.

By neglecting accessibility, news platforms discriminate against people with disabilities. In India alone, there are at least 25 million people with disabilities. As publishers face subscription fatigue and declining audience engagement, addressing accessibility could unlock entirely new segments of readers.

This is why I predict that more newsrooms will turn to AI in the coming year to make their products accessible. AI can address major barriers to accessibility, from reducing manual effort to enabling cost-effective tech solutions. For instance, imagine an AI chatbot trained on the brilliant alt-text resources created by the BBC in a newsroom where desk editors are too swamped to write and add alt text for all images and infographics. Or a tool built on inclusive storytelling guidelines, helping editors save time in checking whether the sources and subjects mentioned in the story reflect the experiences of diverse communities — not just dominant voices in terms of class, caste, gender, and race.

Perhaps we’ll even see an AI anchor that genuinely addresses accessibility needs — like one capable of sign language interpretation.