Every January, media outlets publish their predictions for the year ahead. Most of them are safe bets dressed up as insights: “video will continue to grow,” “AI will be important,” “trust in media will remain low.” Yeah, no kidding.

Here are the predictions nobody’s making but should be, the shifts that will actually matter for how we create and consume media in 2026.

Local News Will Get Weirder

We’re about to see an explosion of AI-assisted local news operations that are technically functional but deeply strange. These won’t be the crude aggregation sites we’ve seen so far. They’ll be sophisticated enough to produce passable coverage of city council meetings, high school sports, and local business news.

The uncanny valley will be real. Stories that look fine at first glance but have subtle tells—awkward phrasings, missing context that a human would catch, quotes that are technically accurate but lack the color a real reporter would include. Readers will struggle to articulate what’s off, but they’ll know something is.

Some of these operations will be transparent about their AI use. Others won’t. The ones that aren’t transparent will get caught, there’ll be scandals, and we’ll have uncomfortable conversations about whether AI-generated local news is better than no local news at all. Spoiler: it’s complicated.

Platform Consolidation Will Accelerate

X (Twitter) will continue its slow-motion collapse, but there won’t be a single replacement. Instead, we’ll see further fragmentation across Threads, Bluesky, Mastodon, and whatever new platforms emerge. This will be terrible for media distribution.

Journalists and publications built their audiences on Twitter over 15 years. That’s gone. Building new audiences across multiple fragmented platforms while maintaining everything else is unsustainable. Many will give up on social distribution entirely, focusing instead on email newsletters and direct relationships.

The death of a unified social media town square will be good for some things—less mob behavior, fewer algorithmic rage spirals—but terrible for breaking news distribution and public conversation. We’ll miss what Twitter used to be while acknowledging it couldn’t stay that way.

The Great Unbundling of Analysis

Organizations focused on AI strategy support and technical domains are already seeing this: the separation of reporting from analysis will become more pronounced. Legacy outlets will lean harder into breaking news and shoe-leather reporting. Analysis, commentary, and explainers will increasingly live elsewhere.

This makes economic sense. Reporting requires institutional resources—legal support, insurance, expense budgets, editorial infrastructure. Analysis requires expertise and writing ability, but one person can do it from anywhere. Why bundle them in the same organization?

We’ll see more subject matter experts building direct-to-audience analysis businesses while relying on traditional outlets for the underlying reporting. This creates dependencies that could be fragile, but it might be more sustainable than expecting one organization to do everything.

Video Will Finally Peak

Hot take: 2026 is the year video content growth stalls. Not declines, but stops being the medium everyone assumes they need to pivot toward. Text and audio will stage quiet comebacks.

Why? Production costs for quality video are high and going higher as audiences expect better production values. Distribution is dominated by platforms that can change rules overnight. And frankly, a lot of content works better in other formats but gets forced into video because that’s where the growth has been.

Smart publishers will diversify away from video dependence, building stronger text and audio operations while maintaining video for content that genuinely benefits from it. The pivot-to-video era will be remembered as a bubble driven by platform incentives that didn’t last.

Paywalls Will Get More Sophisticated (and Annoying)

The era of simple metered paywalls is ending. In 2026, we’ll see dynamic paywalls that adjust based on user behavior, content type, referral source, and predicted conversion likelihood. This will be powered by better data analytics and machine learning.

For users, this means a more confusing experience. Sometimes you’ll hit a wall on the first article. Sometimes you’ll read ten before being blocked. Prices will vary based on what the algorithm thinks you’ll pay. It’ll feel like airline pricing came to journalism.

Publishers will defend this as necessary optimization to maximize revenue. Readers will hate it. But it’s coming anyway because the economics demand it. Expect backlash and calls for regulation around transparent pricing.

Newsletter Fatigue Will Be Real

The inbox is getting crushed under the weight of every journalist, brand, and thought leader launching newsletters. 2026 is when people start aggressively unsubscribing, consolidating their newsletter consumption to maybe five regular reads.

This will hurt everyone except the very best newsletter creators. The long tail of newsletters making modest money will see subscriber declines. Only the truly essential or uniquely valuable will maintain audiences. Many will quit, leading to hand-wringing about whether the newsletter boom was sustainable.

The answer, of course, is that it wasn’t sustainable at the current scale. But the format itself will survive, just with fewer participants and higher quality bars.

AI Detection Will Become an Arms Race

As AI-generated content gets better, the fight to detect it will intensify. Educational institutions, publishers, and platforms will deploy increasingly sophisticated detection tools. AI companies will counter by making their models better at evading detection.

This will be exhausting and mostly futile. Detection will never be 100% accurate. False positives will create new problems—legitimate writing flagged as AI, creators wrongly accused. We’ll need new frameworks for thinking about AI assistance versus AI generation, and those frameworks will be messy and contested.

Media organizations will struggle with policy. Ban all AI use? Require disclosure? Accept it as a tool like spell-check? There won’t be industry consensus, creating confusion for everyone.

The Return of Boring Beats

Here’s an optimistic one: I think we’ll see a small but meaningful resurgence in beat reporting, especially at the state and local level. Not because of altruism, but because there’s commercial opportunity in being the authority on specific domains.

Niche publications covering state houses, regulatory agencies, and local government will find sustainable audiences willing to pay for insider knowledge. These won’t be sexy operations, but they’ll be valuable. The people who actually need to understand what’s happening in state education policy or local zoning will pay for reliable coverage.

This won’t replace everything we’ve lost in local journalism, but it’ll fill some gaps. The model is more trade publication than general interest newspaper, and that’s fine. Sustainability beats nostalgia.

What Actually Matters

Most media predictions are about technology or platforms. What actually matters more is whether we can build sustainable models for journalism that serves public interest, not just audience interest. Whether we can maintain quality and ethics while adapting to new economics. Whether we can rebuild trust that’s been squandered.

2026 won’t solve those problems, but we’ll get clearer on what approaches might work. That’s not sexy, but it’s real. And honestly, real is what media needs more of right now.