What Is Brand Safety and How Do You Ensure It in Programmatic Advertising?
Brand safety is the practice of keeping ads off content and inventory that could damage a brand, verified with third-party tools rather than a single industry standard, since the main cross-industry framework no longer exists.
Global ad fraud losses are projected to exceed $100 billion in 2026, and made-for-advertising sites alone account for roughly 21 percent of programmatic impressions and 15 percent of global programmatic spend with no meaningful return. Brand safety used to have one cross-industry answer to this problem. It no longer does.
Brand safety is the practice of preventing ads from appearing next to content, inventory, or traffic that could damage a brand, verified through third-party measurement tools, exclusion lists, and contextual targeting rather than a single agreed standard. The industry's main shared framework was discontinued in 2024, so brand safety today is assembled from several tools working together, not enforced by one body.
The Old Shared Standard Is Gone
For years, the Global Alliance for Responsible Media (GARM), a World Federation of Advertisers initiative, set the closest thing the industry had to a common brand safety taxonomy. The WFA shut GARM down in August 2024 after X Corp filed an antitrust lawsuit against it and several member brands, and there is no direct successor body setting a single cross-industry standard in its place. Brand safety now runs on the tools underneath GARM, not the framework itself.
What Fills the Gap
Three independent verification vendors, DoubleVerify, Integral Ad Science, and HUMAN Security, now do most of the practical work: pre-bid filtering, post-bid measurement, and invalid-traffic detection layered on top of platform-level tools. A working 2026 brand safety stack combines that third-party verification with exclusion lists (blocklists and allowlists), IAB-level category standards, and configuration specific to each platform, since a blocklist built for the open web does not carry over cleanly to a walled garden or a connected TV environment.
Verification Alone Does Not Solve the Made-for-Advertising Problem
Keyword and category blocklists catch obviously unsafe content, but made-for-advertising sites are built to pass those checks: legitimate-looking pages engineered purely to harvest programmatic impressions. That is why MFA sites still absorb a meaningful share of spend despite years of verification tooling, and why the Association of National Advertisers has put the broader waste from MFA sites, indirect supply paths, and invalid traffic at roughly $22 billion a year. Blocking bad domains is necessary but not sufficient; the inventory also has to be evaluated for whether a real, attentive human is actually there.
How Criterion Global Builds Brand Safety Into Media Plans
On Brookfield's private equity awareness campaign, bot activity across the financial media ecosystem tripled in the weeks before launch, so the plan paired advanced bot verification and attention scoring with contextually rich placements, PE messaging inside PE coverage specifically, rather than broad finance-adjacent inventory, and underperforming inventory was cut in flight as verification thresholds tightened. That combination, contextual targeting plus active verification rather than a blocklist alone, is the same approach behind our broader programmatic advertising and viewability practice, and it is table stakes for the kind of accountable media plan covered in our note on building a paid media program a CFO would sign off on.