In 2021, YouTuber Trevor Jacob deliberately crashed an airplane and filmed the event for a video that eventually drew millions of views. In September 2026, The New York Times used his case to illustrate a broader creator brand safety problem: some creators are escalating what they do because extreme content can outperform routine content in recommendation feeds.
That does not mean every creator is chasing danger, or that every viral post is reckless. It means brands need to look past reach and ask a harder question before they sign someone: what behavior produced that reach? When distribution systems reward surprise, outrage, fear or spectacle, a creator’s growth strategy can become part of a brand’s risk profile.
Table of contents
Jump to each section:
- Why the incentive produces escalation
- Why this lands on brands
- What vetting should actually cover
- The tension brands should name
- What is changing around it
Why the incentive produces escalation
Recommendation feeds create a simple pressure. A routine video competes with everything else a user could watch, while content that triggers a stronger reaction has a better chance of holding attention. The New York Times’ reporting describes creators who interpret that pressure as a reason to keep raising the stakes, even when the next escalation crosses a safety or taste boundary.
The important point for marketers is not that an algorithm explicitly asks creators to take risks. It is that performance can teach creators what to repeat. If the biggest spikes in views consistently come from more confrontational, shocking or risky posts, yesterday’s outlier can become tomorrow’s baseline.
Why this lands on brands
A creator’s audience is usually the reason a brand hires them, but media kits rarely explain how that audience was built. That gap matters because the same instincts that helped a creator break through can show up inside a sponsored idea, especially when the brief asks for something that will “go viral” without defining what is off limits.
The August 2026 fallout around Good Good Golf shows how quickly a creator-led creative decision can become a partner problem. After a promotional video for a co-branded Callaway driver was widely criticized for trivializing violence against women, Associated Press reporting documented Callaway ending the partnership and Good Good stepping away from a PGA Tour title sponsorship. Callaway also acknowledged that it had approved the video, making the lesson larger than creator misconduct alone: brands still own the approval system around creator work.
This is the same accountability problem ContentGrip saw at the US Open, where creator access could not be treated as separate from the brand activation that enabled it. If a creator is there because of their audience, the brand has already entered the content system.
What vetting should actually cover
Follower count, engagement rate and audience demographics are useful, but they do not tell you how a creator behaves when they need the next post to outperform the last one. Vetting should therefore include the creator’s performance history and creative escalation, not just whether their audience looks real.
- Review the back catalogue, not the media kit. Look at the creator’s highest-performing posts and ask what made them work.
- Look for escalation patterns. One old mistake is different from a recent pattern of increasingly risky or provocative content.
- Ask how they plan to make your content perform. If the answer depends on shock, confrontation or an unsafe stunt, treat that as a creative risk before production starts.
- Check enforcement history. Ask about removals, strikes, demonetization or restrictions that could affect campaign delivery.
- Contract for conduct. Define termination rights for dangerous, unlawful or materially brand-damaging behavior, including whether the clause covers non-sponsored posts during the campaign period.
- Decide your position on adjacency. A brand may accept a creator whose personal content is edgier than the campaign. That can be a valid choice, but it should be deliberate.
Dinda Anandita, Account Director at content-led comms agency Content Collision: “If your vetting stops at demographics and engagement rate, you are only checking whether a creator can reach people. You are not checking what behavior taught the algorithm to keep sending people to them.”
The tension brands should name
There is a real tradeoff here. Creators who break through crowded feeds are often good at doing what safer corporate content does not: taking creative risks, moving quickly and making something people cannot ignore. A process that removes every edge can also remove the reason the partnership was valuable.
The goal is not to find a risk-free creator. It is to decide what kind of attention is worth buying and which boundaries cannot be traded for reach. That decision should happen before the creator is selected, not after a post becomes a reputational issue.
What is changing around it
Platforms already draw lines around dangerous content, although the rules are not identical. YouTube’s harmful or dangerous content policy prohibits content that encourages dangerous or illegal activities carrying serious risk of harm or death. TikTok’s current Safety and Civility guidelines restrict dangerous stunts and challenges, while TikTok says another Community Guidelines update will take effect on September 24, 2026. Meta’s standards are broader, while the Times reported that Meta said it prohibits facilitating, organizing or promoting harmful activities to reduce copycat behavior.
Regulators are also starting to target the distribution systems themselves. California’s new rules for under-16s restrict addictive features including algorithmic feeds and autoplay, adding another signal that platform design, not just individual posts, is becoming a policy issue.
For brands, platform enforcement is a backstop, not a vetting system. A creator can stay technically inside platform rules and still be wrong for a brand’s tolerance for safety, taste or adjacency. Reach tells you how many people a creator reaches. It does not tell you what they did to reach them.