Most of what is written AI and marketing this year reads like a warning label. Two new reports released this month demonstrate otherwise: AI is opening up inventory and content opportunities that were previously invisible or undervalued, and there is now data to prove it. One comes from the video advertising side. The other comes from the content side of the creator. Together, they answer a question that Search Engine Journal readers ask all the time: Why does some content take off organically while perfectly competent brand-led work marches on?
The video inventory that no one was buying
The new IAB report for the second quarter of 2026 on AI-powered video results opens with a number that should bother anyone running video campaigns. When Integral Ad Science and Reuters tested the frequency keyword-based brand safety blocking unnecessarily excludes content, they found that 54% of URLs were blocked based on keywords alone, even though the underlying content would be considered appropriate when evaluated for its full context, tone, and intent. More than half. For years, large portions of news video inventory were invisible to advertisers, not because the content was actually dangerous, but because a crude keyword match flagged it.
Multimodal AI is a game changer. Rather than scanning a transcript or headline for trigger words, these tools analyze video, audio, speech, and images together, creating a holistic reading of tone and intent that keyword lists were never designed to capture. I asked Jamie FinsteinVice President of the IAB Media Center, explains directly what this means in practice for media buyers in 2026. Her response was straightforward: “Change always seems like a burden until you realize the cost of not evolving. Teams that don’t revisit their metrics as a result of multimodal AI are going to fall behind.” His specific advice for next week: View your exclusion lists and ask when they were last examined. “For most teams, the answer may have been longer than they would like to admit.”
Timing has a second layer that SEJ readers should note. The 2026 midterm elections are exactly the kind of time when news inventory is historically most excluded, just when public attention is peaking. Finstein’s phrasing on the matter is worth considering: “This concern is understandable, but the math doesn’t support it. Election cycles occur when news consumption is at its peak and public attention is at its highest level. Opting out completely means your brand is absent at precisely the time when consumers are most engaged with media.” The fix doesn’t abandon caution; it’s a question of precision: a report on voter turnout and a partisan commentary do not present the same risk profile, and a content-level assessment can now distinguish them.
Finstein was also blunt about the point that it does not replace human judgment. When asked how marketers should structure monitoring given that AI can still misclassify content in rapidly changing news cycles, she said the priority is transparency and accountability from verification partners, particularly around how edge cases are handled. The opportunity is real, but it’s not a “set it and forget it” upgrade. This is a recalibration that still requires a human to verify the work of the model, particularly live.
For SEJ readers who are also editors, Finstein’s answer regarding content is the most concrete line in the entire interview: “That means making video content easier for review and rating systems to interpret. That starts with clear metadata and transcripts so that each video can be rated individually, rather than relying on broad categories.” The publishers who clean their own metadata and transcripts do the work that allows contextual AI to correctly classify their content as monetizable, instead of leaving it grouped by default into a broad, blocked category.
Content Gap Creators Have Already Solved
The second report presents a completely different angle, but is based on a similar structural vision. Billion Dollar Boya creator marketing agency, partnered with DAIVID’s emotion tracking technology to analyze 5,000 creator-led assets across Instagram and TikTok, mapping what actually drives view rate, engagement, brand favorability and purchase intent against 39 distinct emotional signals. The resulting report, Creator Instinct: Unlock the Social Codeidentifies five specific, measurable behaviors that separate successful content from ignored content, and the gap is bigger than most brand teams probably realize.
The first discovery alone is worth rewriting a content summary. Assets that dominated with the product, benefit, or brand message in the first few seconds saw view rates drop by 44%, brand likeability by 12%, and consideration by 41%, compared to content that created a hook first and let the brand arrive as a reward rather than a pitch. The creators apparently knew this instinctively for years. The data now quantifies exactly what it costs for brands that haven’t caught up.
The second conclusion addresses something that all content marketers have felt but rarely measured: evidence beats claims. Content built around a demo, showing the product in real use, the before and after, the creator’s authentic explanation, outperformed the declarative message “this is amazing” by 33% in brand favorability and 15% in consideration. There’s a helpful quote from designer Laura Adlington in the report that explains why. She explained that showing clothing on one’s own body builds confidence because it allows people to visualize the product in their real life, and that explaining the reasoning behind a style choice builds confidence more than simply stating that something looks good.
The third conclusion is the most useful for content strategists working across categories: there is no better universal emotion. The same emotional register that elevates performance in one vertical actively suppresses it in another. Anxiety dampens beauty and food-related content, but drowns out entertainment, retail, and fashion. Gratitude elevates retail and fashion but stifles beauty and food. This means that a content calendar built around a single brand voice or emotional tone in each category leaves performance on the table by design, not accident.
The fourth and fifth conclusions reinforce each other. Polished, emotionally safe content underperforms. Assets that caused a real, even embarrassing, reaction saw a 25% increase in 2017. organic view rate on safer alternatives, and content combining a raw emotional beat with a positive resolution saw consideration increase by 22% and recommendations increase by 17%. And the ending matters as much as the hook. Content that performed well saw organic view rates increase by 110% across all platforms, and by 318% on TikTok in particular, with engagement on TikTok up 83%. The framework of the report, borrowed from Daniel Kahneman’s Peak-End Rule, is that the audience does not remember entire content. They remember the emotional peak and how it ended. Brands that load their message first and leave the end hanging optimize for the part that viewers forget.
What does it mean if you’re not running paid video or creator budgets
Even SEJ readers who never touch a video purchase or creator contract should read both reports as evidence of the same underlying change. AI tools are getting better to recognize nuance, tone and context rather than simply matching patterns on surface signals, whether it’s a keyword in a video transcript or a generic brand voice applied uniformly across content categories. This change rewards specificity. Video content with clean metadata is correctly classified instead of being globally excluded. Content built around category-specific emotional logic and winning outperforms content built around a single brand model. The common thread running through both reports is the same lesson SEO has been learning all year: tools are getting better and better at telling the difference between truly good content and content that just seems compliant on the surface. This is, for once, good news.
2 steps to follow this week
First, if you’re running or influencing video ad buys, pull your current opt-out lists and brand safety settings and check the date they were last reviewed, as Finstein suggested. If multimodal contextual tools aren’t already part of your review stack, ask your partners what they currently offer and how content is evaluated based on tone, not just topic.
Second, if you’re preparing content for social or creator partnerships, audit your last five briefs based on the particular initial loading issue. If the brand or product appears within the first three seconds of the element, this structural choice alone can cost you almost half of your potential view rate, regardless of the quality of the creative itself. Move the mark to the reward. The data indicates that this is where it works the most.
More resources:
Featured image: Master1305/Shutterstock
