Content themes are a set of topic-level exclusions that Google offers advertisers in Google Ads and Display & Video 360 (DV360), its demand-side platform. An advertiser ticks a theme, such as politics or sensitive news, and Google's classifiers keep the campaign's ads away from videos, websites and apps associated with that subject. The control exists because some lawful content is still wrong for a particular brand: an airline may not want its ads beside coverage of a crash.

As of October 2026, the list in Google's documentation runs to ten themes: Content Suitable for Families, Games (fighting), Games (mature), Health (sensitive), Health (source undetermined), News (recent), News (sensitive), News (source not featured), Politics and Religion. Since October 1, 2026, they have carried more weight, because Google retired most of the older category-based exclusions in DV360 and pointed advertisers to themes instead.

How the control works

Content themes sit inside what Google now calls brand suitability, a group of settings that decides which inventory a campaign may buy. The broadest is the inventory mode, an overall sensitivity tolerance set in three tiers: Maximum, Moderate and Limited. Where the inventory mode adjusts how much profanity, violence or sexual suggestiveness a campaign tolerates, a theme exclusion removes a subject entirely, whatever tier the account uses. The two are designed to be stacked.

In Google Ads, the setting lives in the Content Suitability Center under "Advanced settings", labelled "Excluded content themes". According to Google Ads Help, it applies to YouTube, the Google Display Network and Google video partners, but not to YouTube Sponsorship campaigns. In DV360, the equivalent control sits under "Brand safety" in an advertiser's targeting rules, set once at advertiser level and inherited by every line item.

"News (recent)" covers newly announced news regardless of subject. "News (sensitive)" covers crime, accidents, natural disasters and commentary on controversial social or political issues. "News (source not featured)" removes news from outlets that do not appear on Google News or YouTube News. "Health (source undetermined)", which applies to YouTube only, covers health content from sources Google describes as possibly accurate but less widely cited. Content Suitable for Families includes YouTube videos designated Made for Kids.

The machine-readable version is narrower. The DV360 API exposes the control as the targeting type TARGETING_TYPE_CONTENT_THEME_EXCLUSION, with a ContentTheme enum listing nine values: CONTENT_THEME_FIGHTING_VIDEO_GAMES, CONTENT_THEME_MATURE_GAMES, CONTENT_THEME_NOT_YET_DETERMINED_HEALTH_SOURCES, CONTENT_THEME_NOT_YET_DETERMINED_NEWS_SOURCES, CONTENT_THEME_POLITICS, CONTENT_THEME_RECENT_NEWS, CONTENT_THEME_RELIGION, CONTENT_THEME_UNPLEASANT_HEALTH_CONTENT and CONTENT_THEME_UNPLEASANT_NEWS, plus an "unspecified" placeholder. Families content does not appear in the enum. What buyers see as "sensitive" news is "unpleasant" news in the API.

Classification is Google's own. The advertiser chooses the themes, but Google's systems decide which pages, apps and videos belong to each one, and the buyer sees no theme score per impression. Third-party verification vendors such as Integral Ad Science (IAS), DoubleVerify and Zefr run separate classifiers that can be layered on top through DV360's verification settings, though their categories do not map one to one onto Google's.

Who operates it

On the buy side, content theme settings are usually owned by an agency's programmatic or brand safety team, or by the advertiser's in-house media function. Publishers do not choose which theme their content falls into, and learn of it only indirectly, through lower fill or prices.

Origin and evolution

The idea predates the name. After advertisers paused YouTube spending in 2017 over ads appearing beside extremist videos, the industry built shared vocabularies for unsafe content. The American Association of Advertising Agencies (4A's) formed its Advertiser Protection Bureau on April 10, 2018, and published a Brand Safety Floor and Suitability Framework covering 13 categories on September 21, 2018. The World Federation of Advertisers (WFA) set up the Global Alliance for Responsible Media (GARM) in June 2019, and in September 2020 GARM published a framework of 11 categories, each split into three risk tiers. That framework split brand safety, content no advertiser should fund, from brand suitability, content some advertisers will accept and others will not. Content themes are a suitability tool.

Google's own sequence followed. On October 12, 2022, Marvin Renaud, Director, Global Video Solutions at Google and YouTube, announced the content suitability center in Google Ads, built around three inventory modes plus narrower exclusions. On November 2, 2023, MediaPost reported that Danielle Wolinsky, Google's Global Lead for Advertiser Trust and Responsibility, had presented an "Excluded Themes" option for YouTube and the Display Network at The Brand Safety Summit.

The older controls were then wound down in stages. In September 2024, digital content labels, the film-style maturity ratings DL-G to DL-MA, stopped applying to YouTube ads in every Google buying tool. In 2025, DV360 renamed "Brand safety" to "Brand suitability", said its 17 legacy sensitive categories would be sunset, and named content theme exclusions as one of the controls that would absorb them. On November 7, 2025, Google extended excluded themes to Discover, with YouTube's Home and Watch Next feeds to follow.

Google's DV360 roadmap, last updated May 22, 2026, set an estimated August 2026 date for deprecating category exclusions such as News and Politics and replacing them with advertiser-level content themes. It warned that workflows not moved to the new controls would result in unrestricted serving. The deadline slipped. On August 27, 2026, Google confirmed that most sensitive category exclusions and all digital content label exclusions would end on October 1 across every supported version of the DV360 API. Google's Q3 2026 roadmap says that from that date, inventory modes and content themes cover all YouTube and programmatic DV360 inventory, including display, video, connected TV (CTV) and audio, and that the old signals have been "consolidated into 10 content themes".

Why it matters for marketers

The October change made content themes the main topic-level lever for DV360 buyers. A comparison of the two published lists shows how much the menu shrank. DV360's sensitive categories named Alcohol, Gambling, Tobacco, Drugs, Derogatory, Transportation accidents and Downloads and sharing, among others. None of those has a direct counterpart among the ten themes. Google says their function now lives inside inventory modes, so an advertiser that once excluded gambling with one switch now relies on a tier it cannot inspect by category.

In June 2026, YouTube received Media Rating Council (MRC) brand safety accreditation covering Shorts, its sixth consecutive year of accreditation, with an advertiser safety error rate reported under 1%. That accreditation covers in-stream and Shorts inventory bought through Google Ads and DV360. It does not extend to theme-level classification on third-party programmatic supply, and it does not restore the category-by-category control that some buyers wrote into their policies.

Other platforms have moved in the opposite direction, adding named topics. Amazon added binary content exclusion categories to Amazon DSP on December 18, 2025, across Twitch and third-party inventory in 43 markets.

Limitations and disputes

The central weakness is classification accuracy. A study submitted to arXiv on January 3, 2026, by researchers at the New Jersey Institute of Technology and DeepSee.io examined 3,219 news articles with brand safety ratings from DoubleVerify, IAS and Oracle. According to the paper, IAS rated 53.72% of them unsafe, DoubleVerify 35.27% and Oracle 22.22%. Agreement between IAS and DoubleVerify, measured by Krippendorff's alpha, was 0.19, far below the 0.67 threshold the authors treat as the minimum for tentative conclusions.

News carries the heaviest cost. Three of Google's ten themes are about news, and "News (recent)" excludes breaking stories regardless of subject. Earlier research by Adalytics found that about 30.3% of sampled New York Times articles and 43% of sampled Wall Street Journal articles were labelled unsafe by verification vendors. IAS disputed that analysis, saying the data did not represent advertiser keyword blocklists. Each over-broad exclusion removes demand from lawful reporting.

Transparency is a further dispute. Google does not publish theme-level error rates, the share of inventory each theme removes, or how its "source not featured" judgement is reached, which ties a news outlet's monetisation to its inclusion in Google News. The shared industry reference point has also gone. The WFA discontinued GARM on August 8, 2024, two days after X filed an antitrust suit alleging an advertiser boycott. A judge in the Northern District of Texas dismissed X's claims, and on July 29, 2026, according to MediaPost, the WFA settled with X and agreed not to form or restart GARM or a similar initiative.

What content themes are not

Inventory modes set a single tolerance level across all sensitive content. Content themes remove specific subjects within whatever mode is chosen.

Digital content labels are maturity ratings attached to content, similar to film certificates. They describe how mature content is, not what it is about, and are now deprecated across DV360's programmatic inventory except for content not yet rated.

Sensitive categories are DV360's legacy list of 17 topics. In Google Ads, a separate "excluded sensitive content" list of five, covering tragedy and conflict, sensitive social issues, profanity, sexually suggestive content and sensational and shocking content, still applies to the Display Network and Google video partners as of October 2026.

IAB Content Taxonomy categories are descriptive labels used to classify and target content, not a suitability control. Version 1.0 from 2011 held about 390 categories, and version 2.0 holds 1,180. Outside advertising, "content themes" also means editorial topic clusters in content marketing and search engine optimisation, an unrelated usage.

Recent developments

Third-party vendors are competing to sit on top of the new default. Zefr announced a partnership with Reddit on September 30, 2026, offering multimodal AI classification of conversations rather than broad categories. On October 5, 2026, IAS said it was among the first partners providing brand safety and suitability reporting for ChatGPT ads in a closed pilot.

Timeline

  • April 10, 2018 - The 4A's forms the Advertiser Protection Bureau.
  • September 21, 2018 - The bureau publishes a Brand Safety Floor and Suitability Framework covering 13 categories.
  • June 2019 - The WFA establishes GARM.
  • September 2020 - GARM publishes its framework of 11 categories and three risk tiers.
  • June 2022 - GARM adds misinformation to its framework.
  • October 12, 2022 - Google launches the content suitability center and three inventory modes in Google Ads.
  • November 2, 2023 - Google presents excluded themes for YouTube and the Display Network at The Brand Safety Summit.
  • August 6, 2024 - X files an antitrust suit against the WFA, GARM and several advertisers.
  • August 8, 2024 - The WFA discontinues GARM.
  • September 2024 - Digital content labels stop applying to YouTube ads.
  • 2025 - DV360 renames brand safety to brand suitability and announces the sunset of sensitive categories.
  • November 7, 2025 - Excluded content themes become available for Discover.
  • December 18, 2025 - Amazon DSP adds content exclusion categories in 43 markets.
  • January 3, 2026 - NJIT and DeepSee.io researchers publish a study of vendor disagreement on news classification.
  • May 22, 2026 - DV360 roadmap sets an estimated August 2026 deprecation of category exclusions.
  • June 2026 - YouTube receives MRC brand safety accreditation covering Shorts.
  • July 29, 2026 - The WFA and X settle the GARM litigation.
  • August 27, 2026 - Google announces the October 1 removal of digital content label and most sensitive category exclusions in DV360.
  • October 1, 2026 - Inventory modes and ten content themes cover all YouTube and programmatic DV360 inventory.

Summary

Who: Google offers content themes in Google Ads and Display & Video 360; advertisers, agencies and in-house media teams set them, while publishers and creators are affected by the exclusions without choosing their own classification.

What: A list of ten topic-level exclusions, covering news, health, politics, religion, mature and fighting games and family content, that keep ads off content Google's classifiers associate with each theme.

When: Google presented excluded themes in November 2023, extended them to Discover in November 2025, and made them the main topic control for DV360 programmatic inventory on October 1, 2026, when most sensitive category exclusions ended.

Where: YouTube, the Google Display Network, Google video partners, Discover and Demand Gen, and from October 2026 all programmatic DV360 inventory, including display, video, CTV and audio.

Why: Advertisers want to avoid lawful content that conflicts with their brand. Content themes provide that at topic level, but contested classification accuracy, the blocking of legitimate news, and the loss of finer category controls limit how precisely they work.