The Newsroom That Publishes a Fake First Doesn't Get a Second Chance
Why trust checks belong inside the story workflow—not bolted on around it
Protecting your reputation in the age of deepfakes and questionable sources doesn’t come down to having the strictest verification policies. It comes down to an architecture decision: Whether verification is built into the workflow itself, or if it’s an extra step that deadline pressure can delete. Most newsrooms have built their architecture wrong; Avid for News helps you build it right.
The threat of misinformation
Most of your audience no longer comes to you for your journalism. They encounter it somewhere else first: In a feed, sandwiched between other videos, stripped of the context your newsroom wrapped around it.
For the first time, more people worldwide get their online news from social media and video networks than from news sites and apps1. And these are the environments people trust least: The Reuters Institute's 2026 Digital News Report puts trust in news on social media at 22% against 37% for news overall2.
Your audience now meets your reporting where their guard is highest and your margin for error is lowest. Fabrication has always been able to outrun verification; now it does so on the channels that have become the front door to your journalism.
There's a phrase going around the industry: Fake is about to get real. It already is. Generative AI has collapsed the cost of creating convincing fake material to near zero: A plausible image in seconds, a cloned voice from a one-minute sample, fabricated quotes and manipulated video on demand. And concern is climbing with it: 62%3 of people worldwide now say they worry about what's real and what's fake in online news, up sharply this year, with the rise driven explicitly by fear of deepfakes and AI-generated content. In the US, it's nearly three in four.
This isn’t a niche anxiety or a problem for tomorrow—it’s happening everywhere right now. The World Economic Forum ranks misinformation and disinformation as the top global risk over the next two years4, for the second year running, ahead of extreme weather and armed conflict.
Meanwhile, the ground beneath every newsroom is eroding: Global trust in news has fallen to 37%5, the lowest level on record, down across most of the markets measured. Much of that decline is about the wider information environment, not any single newsroom. But it sets the stakes. Trust isn't a reserve you can draw down. It's a one-mistake asset. The newsroom that publishes a fake isn't graded on its average—it's remembered for its worst day.
The policy illusion
Most news organizations believe they have this covered. They have verification policies, editorial guidelines, training programs, and a document somewhere that says all user-generated content must be checked before broadcast.
Under deadline pressure, policy loses to workflow.
When a journalist is twenty minutes from air with a breaking story, any verification step that lives outside their production system (e.g. a separate browser tab, an external tool, another team's queue) is a step that can be skipped. Not because journalists are careless, but because the entire gravitational pull of a breaking news operation is toward publish. A process that depends on people interrupting their own momentum, repeatedly, under pressure, isn't a process. It's a hope.
The newsrooms getting this right have stopped relying on discipline and started relying on design. The check now happens where the work happens. There is no separate step to skip.
Verification is a story problem, not a journalist problem
There's a second structural failure in how most newsrooms approach verification, and it comes from tool-centric thinking.
In a traditional setup, each desk works in its own system: The broadcast team in the rundown, the digital team in the CMS, the social team in their publishing tools. When a questionable asset enters the building, each team meets and checks it separately or assumes another team already has. Verification gets duplicated where it's done well and silently skipped where everyone thought someone else owned it.
That gap used to sit at the edges of the operation. It now sits at the center. As audiences move to social and video platforms, the social and digital desks are no longer a side channel pushing out clips after the broadcast. They are how most of your audience meets your journalism in the first place. They work fastest, publish to the least forgiving environment, and are exactly where a verification step performed somewhere else gets lost. The desk under the most speed pressure is the one your reputation now most depends on.
The story-centric model fixes this at the root. When every contributor works from a single shared story, verification becomes a property of the story itself. A source checked once is checked for everyone. An asset flagged as suspect is flagged for everyone. The work of establishing trust travels with the story through every stage of production and out to every platform: broadcast, web, and social alike.
This is what the industry conversation about deepfakes keeps missing. The answer to industrial-scale fabrication isn't heroic individual vigilance. It's an operating model where trust is established once, at the story level, and inherited by every output.
Putting this into practice with Wolftech News and MediaCentral
A fake claim spreading online does damage in seconds, not hours. The tools journalists use must work at the speed of the threat: In real time, inside the conversation as it happens, not in a review queue behind it. Wolftech News is built on that requirement and the assumption that a fake is already at the gate.
As they write, your journalists check claims in real time against credible global sources, powered by Factiverse, without leaving the story. Verification isn't a research task they switch to; it becomes part of writing.
And it has a second effect newsrooms notice fast: it saves time. Checking claims in text or video inside the workflow takes a fraction of the time of manual research, with its separate tabs, searches, and cross-references. The check that protects the story also speeds it up.
Video is where this matters most. More than three-quarters of people now watch online news video each week6, most of it on third-party platforms rather than your own. And video is the format where synthetic manipulation is hardest to catch by eye.
When a questionable clip is uploaded to Wolftech News, your team screens it for synthetic manipulation—deepfaked video, cloned audio, generated imagery—with best-in-class detection built directly into the story workspace. The clip is checked long before it reaches a rundown or a publish queue. Nobody needs to know which model to use, where to upload a file, or whose job the check was. The capability sits exactly where the work happens.
C2PA content provenance support extends the same principle beyond your newsroom, toward an industry standard where content carries its origin and edit history with it, so your audience and your partners can trust what leaves your building as much as you trust what's in it.
The publish decision still belongs to your editor, as it always should. What the workflow guarantees is that the question “can we trust this?” gets asked systematically, when it matters, by default rather than by exception.
The decision in front of every news leader
The volume of synthetic content will not plateau. The tools to create it are improving exponentially, and the incentive of bad actors to target newsrooms only grows because of the power a fake is given when it’s amplified by a trusted outlet.
Audiences are already walking away. 42%7 now avoid the news at least some of the time, up from 29% in 2017, and every time an unverified fake reaches air, more will join them.
The question facing every organization’s head of news and chief engineer isn't whether to take verification seriously. You already do. The question is whether it actually happens and where: Inside your workflow, where it survives the worst, fastest, most dangerous day, or alongside it, where it gets skipped precisely when it matters most.
Your audience won't see the difference in your architecture. They'll see it in your output, and they'll judge you on the fake you didn't catch. Build the checks into the story, because trust, once published away, doesn't come back.
If your verification process lives outside your workflow, it will be skipped when it matters most. Request an Avid for News demo to see how it should be done.
Sources
In an article about trusting sources, it’s important we cite the ones we used!
1. “Social media and video networks are, for the first time, the single most widely used way of accessing online news (54%), ahead of news organisations' own websites and apps (51%).” Reuters Institute for the Study of Journalism, Digital News Report 2026, University of Oxford, June 2026.
2. “Trust in news on social media stands at 22% against 37% for news overall.” Reuters Institute for the Study of Journalism, Digital News Report 2026, University of Oxford, June 2026.
3. “62% of people worldwide are concerned about what is real and what is fake online when it comes to news, up 4 percentage points this year.” Reuters Institute for the Study of Journalism, Digital News Report 2026, University of Oxford, June 2026.
4. “Misinformation and disinformation ranked the top global risk over the next two years, for the second year running.” World Economic Forum, Global Risks Report 2025, January 2025.
5. “Global trust in news has fallen to 37%, the lowest level recorded since measurement began in 2015.” Reuters Institute for the Study of Journalism, Digital News Report 2026, University of Oxford, June 2026.
6. “More than three-quarters of people (77%) globally consume online news video each week, with growth concentrated on third-party platforms.” Reuters Institute for the Study of Journalism, Digital News Report 2026, University of Oxford, June 2026.
7. “42% of people now avoid the news at least sometimes, up from 29% in 2017.” Reuters Institute for the Study of Journalism, Digital News Report 2026, University of Oxford, June 2026.