Overblocking the news isn't brand safety. It's a performance leak.
New performance data, a room full of media executives who rarely agree on anything, and a stack of antitrust settlements are turning journalism support from an ethics argument into a math problem.
By Chris Marine, Founder and CEO, Campfire
For years, the case for advertising against news content sounded like a values pitch. Support journalism because democracy needs it. Fund reporters because someone has to pay for the truth. It's a good argument. It's also an easy one to nod along to in a budget review and then quietly deprioritize when the actual media plan gets built, because values pitches rarely have a line item attached.
That's starting to change. Not because anyone got more virtuous. Because the numbers caught up.
The data
On July 31, The Trade Desk published new testing on News Navigator, the tool it built with news rating partners Ad Fontes Media and Mobian to evaluate credibility at the page level instead of blocking entire categories with keyword lists. In aggregate tests, campaigns that used the tool to shift spend toward highly rated journalism increased their investment in that content by 24% and saw an 8% drop in cost per acquisition and a 9% lift in click-through rate.
It's worth saying plainly that this is The Trade Desk's own aggregate testing, not an independently audited study, and it's being published on a page built to sell News Navigator. Read it with the same skepticism you'd apply to any vendor claiming their own tool works. But the reason it's worth your attention isn't the study. It's who showed up to agree with it.
We'd add ourselves to that list. Not because it makes for a better pitch, but because it matches what we keep seeing in our own media plans. Across national brands and regional challengers alike, local journalism consistently outperforms the national environments most plans default to. Banners running against local news coverage routinely outpace the same creative running on national sites, and by a wide enough margin that it changes how we build a plan. That gap doesn't shrink for the big clients or grow for the small ones. The audience is there locally too, and it pays attention.
The consensus
The data came packaged with commentary from a genuinely unusual lineup. TIME, Bloomberg Media, General Motors, and The Trade Desk are all making some version of the same argument in public right now. Brand-safe does not mean news-free.
Mark Howard, TIME's chief operating officer, pointed to a specific failure of the old model. The outlet's Person of the Year coverage of Taylor Swift got flagged as unsafe by legacy verification tools, simply because the story touched on feminism and the exclusion of women from power. That's not a brand safety problem. That's a keyword filter mistaking a topic for a threat.
Shenan Reed, General Motors' chief global media officer, takes a more useful approach. GM built an internal framework, partly on Ad Fontes Media's data, that defines where the brand is comfortable showing up rather than avoiding an entire category. Her read on the audience sitting inside quality news content was blunt: "The audience is there." Reach was never the problem. Tooling was.
Jeff Green, founder and CEO of The Trade Desk, framed the shift as a reasoning problem more than a technology problem. The goal is to identify the news environments that are factual and less polarizing, instead of treating "news" as one undifferentiated risk category.
And Vanessa Otero, founder and CEO of Ad Fontes Media, connected it to something bigger than any single campaign. As AI makes cheap content trivial to produce, human reporting, the expensive, fact-checked, legally reviewed kind, gets more valuable, not less. That list includes investigative work, local coverage, and war and disaster reporting, none of which gets replaced by a language model, and all of which keyword blocklists have been quietly starving for the better part of a decade.
The part that makes this durable
This is what turns a good week of trade press into an actual shift. The ground underneath collective brand safety standards has been getting less stable for two years, and it just gave way.
In April, the FTC and a coalition of state attorneys general settled with Dentsu, WPP, and Publicis over allegations that the holding companies coordinated shared "brand safety" standards through industry trade groups rather than setting them independently. The settlement bars those agencies from colluding on unified exclusion standards going forward. Whatever you think of the politics wrapped around that case, the mechanism matters well beyond the three named. Regulators have now said, in an enforceable order, that agencies coordinating a shared blocklist with their rivals is an antitrust problem, not just a service they offer clients.
Then in late July, X and the World Federation of Advertisers settled the lawsuit that had already forced the shutdown of the Global Alliance for Responsible Media back in 2024. GARM is not coming back. Neither is the idea of one industry-wide standard that every agency plugs into and calls it a day.
Put those two developments next to the News Navigator data and the signal points the same direction. The era of "we all use the same blocklist" is over, both because it doesn't perform and because it's now legally risky. That leaves brands needing a framework built and defended for them specifically, not a standard borrowed wholesale from an industry group.
The part nobody's saying
Everything above puts the correction on the demand side, on advertisers and the agencies buying for them. That's only half the story, and it's the easier half to tell, because it doesn't implicate the publishers and networks the industry keeps asking advertisers to trust more.
For the better part of the last fifteen years, plenty of publishers and networks chased the same audience-extension model everyone else did. Instead of building and defending the value of their own inventory, they started reselling impressions across other properties entirely, packaged and priced on audience targeting parameters instead of the publisher's own journalism. Functionally, that's an agency's business model wearing a publisher's name. And that’s how they’d want you to perceive them.
The trouble is that a publisher operating that way doesn't answer to the advertiser the way an agency is supposed to. It answers to its own margin and its own shareholders. There's no fiduciary obligation running toward the advertiser, just a sales relationship dressed up as a media partnership. Some of the holding companies got exactly this kind of scrutiny this year, in the FTC settlement above. Publishers and networks running the same play haven't gotten nearly the same attention, and they should.
That gap matters here specifically because it's the same story from the other side. Newsrooms didn't only get starved by overcautious blocklists on the buy side. Plenty of them got starved on the supply side too, by parent companies choosing audience arbitrage over reinvesting in the journalism that built the audience worth extending in the first place.
I saw a smaller version of this up close a long time before I had language for any of it. Nearly two decades ago I was starting out at a Tegna owned NBC affiliate as one of their first multimedia journalists, and I asked a producer, more naively than I realized at the time, why our newsroom (the #1 with audience share in the market by a long shot) kept losing staff and resources every single year. He told me if I actually wanted the answer, I needed to stop looking at the newsroom and start looking at where the money came from. Then he pointed straight up toward the ad sales floor above us and said the real place to make a difference wasn't down on the assignment desk, it was up there. That one offhand comment is the reason I do this for a living. Years later, having served on the ad sales side and now the buy side I'm still chasing the same question, just at a bigger scale and through different pipes instead of a sales floor.
Where this leaves marketers
The easy fix is also the wrong one. Swapping an old vendor blocklist for a new vendor score isn't the same as solving it. Ad Fontes Media and Mobian are good at what they do, but outsourcing your brand's judgment to a single rating system is the same mistake with better branding.
A real framework pulls together a few things no single vendor sells as a bundle.
A financial services brand and a beverage brand shouldn't run the same suitability settings. What counts as risk depends on who you are and who you're trying to reach.
Page-level evaluation instead of domain-level blocking. The Trade Desk's own example makes the case. The problem was never TIME, it was a filter that couldn't tell a Person of the Year story from an actual threat.
Direct publisher relationships matter here too. Rating data tells you what's credible, but it doesn't replace knowing your publishers, how they're funded, and what terms actually support the journalism you're trying to reach.
Know your supply path. Understand how many hands your money passes through before it reaches the newsroom you're claiming to support.
And measure it. Quality news doesn't get a pass on performance. It has to earn its place in the plan like anything else, which is what makes this argument different from the old charity pitch.
That combination is harder to build than turning on a single vendor's toggle. It's also the only version of this that isn't exposed to the kind of collusion risk regulators just spent a settlement unwinding. It's your brand's judgment, applied consistently, and defensible on its own terms.
The actual argument
Supporting journalism was never a charitable line item, even when it got pitched that way. What changed is that the cost of getting it wrong finally became visible, showing up in lower click-through rates, in wasted spend on made-for-advertising junk, and now in regulatory settlements that treat shared exclusion lists as a competition problem rather than a best practice.
The advertisers who figure this out first won't be the ones with the strongest values statement. They'll be the ones who did the harder work of building their own framework instead of renting someone else's.
