The Year AI Messed With Our Heads

In partnership with

If you work with information for a living right now, you probably have two tabs open at all times. One is whatever you're working on. The other is a quiet existential crisis you keep minimising.

2025 was the year AI stopped being a "coming soon" headline and became ambient. It seeped into everything — workflows, pitches, propaganda, policies, and (more awkwardly) our relationships.

A 28-year-old entrepreneur told me he uses ChatGPT to reply to emotional friends because he's scared of "saying the wrong words." A 39-year-old I spoke to on a dating app confessed that he was feeding text exchanges into an AI tool to figure out what to do next. "It told me not to double text," he said, like the machine had issued a court order.

The more I heard these stories, the more unsettled I felt. If people are now using AI to buffer the discomfort of being human in real time — especially in moments that require empathy, intuition, awkwardness — what does that mean for journalism? For a profession built entirely on feeling more, on noticing what's off, on sitting with silences, on recognising the quote that cracks because it's clumsy, not because it's clean?

So I decided my last newsletter of the year should try to make sense of how we dealt with AI in 2025 — as journalists trying to do our jobs, as people trying to make sense of a world that's suddenly more automated and somehow lonelier. And perhaps, what we might carry into 2026.

Media turned to machines 

Indian media houses used AI quite visibly. Aaj Tak introduced Sana in 2023 and DD Kisan launched AI Krish and AI Bhoomi in May 2024 — tireless, always-on virtual presenters built for 24x7 news bulletins. They got more popular even as journalists raised questions about labour displacement and credibility.

But alongside the novelty came the mess. Media publications were called out for publishing articles that appeared to be copied verbatim from ChatGPT. The Economic Times published an AI-generated story about Trump's fake Obama arrest that invented a quote from disinformation expert Nina Jankowicz. She exposed what she called the "AI ouroboros" — her fabricated words looping into other AIs, eroding trust in Indian newsrooms chasing global ad revenue.

In my experience, giant-sized mess happens because newsrooms are under-resourced, underpaid, and under pressure to publish constantly. So when AI arrived promising speed, polish, and relief, it felt like the perfect unpaid intern who never sleeps. But it also showed up like a pickpocket in a crowd.

Globally, journalists are racing to adopt AI. A Thomson Reuters Foundation survey of over 200 journalists from the Global South found that over 80% use AI tools in their work, with nearly half integrating them daily. Meanwhile, the Reuters Institute's 2025 report found only 12% of the public are comfortable with fully AI-made news — rising to just 21% with human oversight.

So, the takeaway: journalists are using AI more than ever, while audiences trust it less than ever.

The inbox got shinier, the reporting got vaguer

AI became the tool of choice for reporters to send cleaner pitches, faster. Sometimes that's genuinely useful — clearer framing, tighter writing, fewer rambling emails.

But it also created a new species of problem: the pitch that looks perfect and tells you nothing.

Over the last year, I've read drafts where I could see the symmetry: polished paragraphs, confident tone, barely any reporting. I even went on LinkedIn to see if other editors felt the same, and the response was telling: many did.

Because AI doesn't know what matters in a story. It doesn't know which quote to keep — the one that reads well, or the one that sounds clumsy but carries emotion. It doesn't know that the colour of someone's shirt, the way they fidget, or how their home smells can carry as much weight as a statistic.

AI can make writing look like reporting. Which means editors now have a new job title: human lie detector.

And the part we don't say loudly enough: AI adoption in Indian journalism is wildly unequal. Some newsrooms have tool budgets, training, and legal oversight. Others are filing three stories a day, chasing unpaid invoices, managing basic access issues. In that world, AI isn't a democratising force. It's another divider.

And still, AI made some parts of the work easier

I'll admit something that would've made 2019-me deeply suspicious. I am a little less scared of long documents now.

This year, one of the most practical things I've used is Google's NotebookLM — the one genuinely useful tool that lets you upload sources and interrogate them. Because it's trained on your documents (not the internet's collective fever dream), it actually helps with the parts of journalism that quietly destroy your soul: policy drafts, legal annexures, 300-page PDFs that seem designed to make you abandon curiosity.

Journalists are using a range of AI tools. IJNet recommends Trint for transcription, Poynter suggests Otter.ai combined with ChatGPT for summarising, and reporters praise Pinpoint for document dumps and Claude for summarising with fewer hallucinations.

Used carefully, AI didn't replace thinking — it protected time for it. And in a country where power increasingly hides behind paperwork, that matters.

When pranks turned into panic

But AI also means messy. While some of us were learning to use AI responsibly, others were learning just how easy it is to create chaos.

In October, a 22-year-old journalism student in Lucknow used ChatGPT to add a leopard to a selfie taken on his balcony. He sent it to a WhatsApp group as a joke. Within hours, local police and forest officials were at his door after frantic calls from residents about a leopard on the loose.

The boy was let off with a warning. But the damage was done. Over the next days, more AI-generated leopard images spread across Lucknow's Rajni Khand, Ruchi Khand, and Ashiana neighbourhoods. Forest department officials had to scan hours of CCTV footage, create patrol teams working in shifts, and calm panicked residents.

All six or seven images that went viral turned out to be hoaxes. Forest officials eventually started using ChatGPT and other AI tools themselves to analyse the fakes — noticing pixels were too large, that the images looked like stock photos pasted onto local scenery.

It's darkly funny: we've reached the point where forest officials need AI to detect AI-generated leopards that were never there.

And while students were pranking their neighbourhoods, politicians were playing a far more dangerous game.

Archis Chowdhury's investigation for BOOM revealed how generative AI is being used to mass-produce cheap, fast, cinematic propaganda — content engineered to feel like proof, arriving when official information is scarce, filling the vacuum with high-emotion narratives.

This is the crucial shift: propaganda isn't only text anymore. It's atmosphere. It's mood. It's cinema. And cinema travels in India — especially when it arrives on WhatsApp without watermarks, without disclosure, with the confidence of a trailer.

Editors are now stuck with a brutal asymmetry: journalism is slow because truth is slow. Propaganda is fast because it doesn't need to be true.

The money problem nobody solved

Let's talk numbers, because they're absurd and crucial. Digital media now commands around 46% of India's ₹1 lakh crore advertising market, according to Crisil Intelligence. India's M&E sector crossed ₹2.5 trillion in 2024, per the FICCI–EY report, with digital continuing to grow.

And journalism? Survival is the only benchmark of success. Ad money doesn't reliably reach independent outlets. Platform distribution is fickle at best, hostile at worst. 

Nieman Lab's predictions for 2026 are blunt: AI will not rescue journalism's business model. There will be no meaningful licensing windfall. The economics simply don't add up.

So what survives? Experts are betting on these, and I concur:

Original reporting. AI can rewrite. It can summarise. It cannot build trust over weeks or spot the tiny contradiction that cracks a story open.

Direct relationships. As platforms become unreliable landlords, Nieman predicts more media businesses will become "events businesses" — journalism that builds community through newsletters, memberships, gatherings has a fighting chance.

Editorial courage. Deciding what deserves time. Deciding not to chase the cheapest traffic. Deciding to publish the story that won't trend today but will matter six months from now.

I've watched Indian journalism shrink and mutate over the last decade— newsrooms collapsing, desks thinning, survival repackaged as "pivoting." 2025 didn't fix that. It amplified it.

But it also clarified something I find stubbornly hopeful: as the internet becomes more synthetic, real information becomes more valuable. And if AI makes content easier to produce, it also makes the difference between content and journalism easier to see.

I'm not a tarot card reader, but I know that in 2026, the outlets that survive won't be the ones that learned AI the fastest. They'll be the ones that used AI to protect time for reporting — and then defended that reporting with everything they had. The only thing that will save journalism is the humans who refuse to let it die.

For this issue, I asked my colleagues at BOOM to share their favourite read or watch of the year. Not “best” or “important”, just the one that stayed with them. Here’s the list (In their words).

Got a story to share or something interesting from your social media feed? Drop me a line, and I might highlight it in my next newsletter.

See you in your inbox, every other Wednesday at 12 pm!

Turn AI into Your Income Engine

Ready to transform artificial intelligence from a buzzword into your personal revenue generator?

HubSpot’s groundbreaking guide "200+ AI-Powered Income Ideas" is your gateway to financial innovation in the digital age.

Inside you'll discover:

• A curated collection of 200+ profitable opportunities spanning content creation, e-commerce, gaming, and emerging digital markets—each vetted for real-world potential

• Step-by-step implementation guides designed for beginners, making AI accessible regardless of your technical background

• Cutting-edge strategies aligned with current market trends, ensuring your ventures stay ahead of the curve

Download your guide today and unlock a future where artificial intelligence powers your success. Your next income stream is waiting.

Get Your Guide

Was this email forwarded to you? Subscribe