Recommendation algorithms route reach toward content that produces surprise, conflict or spectacle, and under that structure a creator's own growth strategy migrates onto your brand's risk sheet. What to check before signing isn't the media kit — it's the highest-performing content in the back catalogue — and you need to separate one old mistake from a recent pattern of enforcement actions, name the conduct explicitly in the contract if a creator's pitch leans on shock or an unverified stunt, and decide before the campaign starts who calls a pause if the risk profile changes mid-flight.
Why do algorithms push creators toward riskier content?
In September 2026, the marketing outlet ContentGrip pointed to a case The New York Times had used to make this argument: YouTuber Trevor Jacob deliberately crashed his own small plane for views in 2021 and filmed the entire stunt. ContentGrip's framing was that when a recommendation system rewards surprise, outrage, fear or spectacle, a creator's growth strategy itself becomes part of a brand's risk profile. Feeds rewarding attention-grabbing content isn't new, but the incentive it creates for an individual creator is specific: a format that performed last month often has to be pushed one notch further to hold the same reach this month, and that pressure builds entirely outside your contract, which means you can't pre-empt it from the brief. The question isn't what a creator once said wrong — it's which direction their reach graph keeps getting pushed.
Why check the back catalogue instead of the media kit?
A media kit is a creator's own highlight reel, and a quick scroll through recent posts mostly re-confirms the same highlights. What you actually need to see is which posts drove the most views and shares, and how the production behind those posts has changed as you move closer to today. Scanning whether an audience looks real is a different exercise from scanning a performance history for an escalation pattern. This is also a different question from a pre-signing audit of old posts for statements that would clash with your brand today — that audit looks for something that was once a problem; this one looks at which direction the account is currently moving. Reading this alongside our guide to auditing a creator's past posts before you sign makes clear the two checks fill different gaps.
How do you tell one mistake from a pattern?
An old mistake and a recent pattern of increasingly risky content are different signals, and the most direct way to catch the latter is to ask about enforcement history. Find out whether the creator has had posts removed, platform strikes, demonetization or reach restrictions recently — and if so, how many times and how recently. An enforcement history doesn't automatically disqualify a creator, but it does mean that if the account gets restricted again for the same reason mid-campaign, your reach can disappear for a reason you never controlled. Put this question before the deal is confirmed, not after the media kit stage, so you're not learning about it once the campaign is already live.
Why does your contract need to name the conduct?
If a creator's pitch for why a piece of content will perform rests on confrontation, shock or an unverified stunt, that's a creative risk to deal with before production starts, not a problem to patch after the footage exists. A generic brand-safety clause (no illegal, hateful or sexual content) is probably already in your contract, but it rarely names "views driven by a dangerous stunt" as its own category. Add a pre-approval step for the content direction at the brief stage, require separate written sign-off for any shoot involving physical risk, and give your brand a standalone right to re-review content if a creator's recent output has visibly escalated in risk during the contract term.
Who decides when risk changes mid-campaign?
A creator who was low-risk at signing can pivot toward riskier content mid-campaign, and if nobody has already decided who calls it and what happens next, that decision gets made improvised, under pressure. Set the conditions that auto-pause a campaign (a repeat enforcement action, a public risky-stunt post), name who gets the legal alert, and name who owns the call — all before the campaign launches, not during a crisis. Example: on a 3-month campaign with a $23,000 budget, if 60% ($13,800) is already spent when a creator draws a platform enforcement action for a risky stunt, a brand with no pre-set clause has to decide whether to keep spending the remaining 40% ($9,200) in the middle of the crisis — a brand with the clause just executes a standard it already agreed to. Pairing this with our influencer contract checklist lets you close this decision before the campaign starts. None of this tells you whether a creator is worth signing on performance — that's a separate question, and one Hyperstar is built to answer by matching creators on actual revenue contribution instead of reach. If you're tightening up your brand safety process this quarter, get started.
