Advertisers try to scrub ads from Breitbart and other sites, resorting to double checking where their ads get placed by automated buying systems
Jessica Toonkel / Reuters : Tweets: @darkecologist and @profcarroll Tweets: @darkecologist : @HarmfulOpinions http://www.reuters.com/... seems like a mistake. once you start curating ads, people will think (even more) that you endorse David Carroll / @profcarroll : A newly populist polarization means programmatic ad buying now a burden on brands. A flawed model. http://reut.rs/2nMQSzU via @Pagankennedypic.twitter.com/ 2gOpBGWtfK
Context & Ripple Effects
Programmatic buying was sold as reach without responsibility: algorithms place brand ads across the open web, and nobody at the advertiser signs off on each page. That bargain is breaking — advertisers are now manually auditing where automated systems put them, with Breitbart the emblematic destination they want out of, and David Carroll framing the whole apparatus as 'a flawed model' that burdens brands with endorsements they never chose.
The timing matters: within days, an analysis found big-brand ads running beside politically extreme and derogatory content across the web (ads for major brands placed next to extreme content), confirming this wasn't one site's problem but a structural feature of automated supply.
First-order effects
- Advertisers pulling placements from Breitbart directly cut the site's programmatic revenue, and every advertiser that follows has to absorb new labor costs from manually verifying where its ads land.
- Publishers on the open web face a credibility tax: even non-extreme outlets inherit the burden of proving their inventory is brand-safe when buyers can no longer trust the auction.
Second-order effects
- Money migrates toward controllable channels — consistent with companies already paying Facebook to promote positive press rather than fund open-web publishers (ad dollars steering toward Facebook promotion) — concentrating spend in walled gardens that promise curated adjacency.
- Verification and whitelisting become a required layer of the stack, and platforms' own moderation gets pulled into the fight, as later seen when Facebook's mix of AI and human review kept snaring legitimate news alongside the bad content it was hired to catch (Facebook's ad-screening snares legitimate news).
Third-order effects
- If brand-safety scrutiny combines with the growing argument that ad effectiveness itself is inflated by benchmarks counting people who would have bought anyway (skeptics say online ad effectiveness is overhyped), the open-web programmatic model faces pressure on both price and proof — pushing the industry toward fewer, larger intermediaries who can guarantee both placement and measurement.
- The deeper shift is accountability migrating back onto humans: curation was supposed to be automated away, but polarization made adjacency a brand decision, meaning the industry either rebuilds human oversight into buying or accepts reputational risk as a line item.
The trend: Automated advertising is being forced to reinsert human judgment — first for brand safety around polarizing publishers, and increasingly for whether the ads work at all.