How Social Media Algorithms Profit from Dangerous Driving Content
BBC Verify has confirmed what the engagement economy has been quietly monetizing for years: dozens of TikTok videos depicting high-speed car chases, footpath driving and wrong-side-of-the-road racing remain easily searchable on the platform.
Algorithmic reward, fatal outcome
At least one of the young people killed in last week's Middlesbrough crash appears in multiple such clips. The algorithm did exactly what it was optimized to do. It surfaced them until external pressure forced a takedown.
What the data shows
The BBC investigation identified a content category — call it "driving stunt" content — that consistently clears TikTok's moderation thresholds while still generating measurable watch time. Cleveland Police have launched a wide-scale investigation into the criminality preceding the fatal crash, though they note there is no evidence the filming was conducted specifically for TikTok. The distinction is legally relevant but commercially irrelevant: the platform captured the engagement either way.
TikTok's stated policy prohibits content that "could lead to physical harm." The company removed videos flagged by BBC Verify after publication. The lag between upload, algorithmic distribution, and removal is where the monetization happens — creators build audiences, brands book sponsorships, and the platform sells attention against the metric of completed views.
The structural read
Treating creators strictly as media businesses, the dangerous-driving vertical is a textbook case of algorithmic decay risk. A creator's follower count and view totals reward volume and virality, not safety compliance. The same popularity signal that attracts sponsorship dollars can be retroactively reclassified as harmful once a news cycle hits. Conversion metrics built on that audience evaporate overnight. A parallel logic applies to copy trading, where follower count and social proof routinely mislead retail capital into bad bets — both markets treat popularity as a quality signal it almost never is.
The bottom line: the Middlesbrough case is not a moderation failure. It is a moderation system functioning as designed, with a fatal externality that only becomes newsworthy when it reaches the front pages.
