The majority of traffic from X may have been fake during the Super Bowl

Original Article ↗ Hacker News Discussion ↗

Reports that a bot-detection firm found unusually high levels of fake traffic from X (formerly Twitter) during the Super Bowl prompt broader concerns about spam, bots and ad fraud on the platform. Commenters describe timelines overrun with sex bots, crypto scams and engagement-farming blue-check accounts, and some advertisers report surging cheap clicks that no longer convert. Others question the reliability and methodology of the bot estimates, but there is wide agreement that X’s anti-spam controls and API restrictions have not delivered on promises to reduce inauthentic activity.

Scope and reliability of the CHEQ claims

  • Several commenters are skeptical of CHEQ’s data:
    • Relative traffic numbers across platforms (TikTok vs Facebook vs X vs Instagram) look odd and “marketing‑like.”
    • Ad-fraud tools are described as largely blind inside “walled gardens” and dependent on platform APIs and “trust me” data.
    • Methodology based on post-click behavior may only capture bots that follow links, missing many that don’t.
  • Others think X’s bot problem is clearly severe, but doubt it is quite as extreme as CHEQ implies.
  • Overall, people see the report as suggestive at best, not definitive.

User experience on X/Twitter

  • Many report X is flooded with:
    • Sex bots, OnlyFans catfishes, crypto scams, “watch live” sports scam links, and engagement-farming replies.
    • Paid blue-check accounts posting low-effort or AI-generated content to game the algorithm.
  • Some say the platform has always been a “wasteland”; others see a distinct worsening since the acquisition.
  • A minority report similar or improved experience (cleaner feeds, fewer crypto scams) and attribute negativity to bias.

Blue checks, incentives, and spam economics

  • Paid verification is widely seen as having:
    • Created an “engagement tier” where paying accounts’ replies are boosted.
    • Reduced incentives for the platform to ban paying spammers.
    • Enabled more effective impersonation (including high-tier gold checks) for scams.
  • Speculation that bots use stolen or virtual cards to subscribe.
  • Ad-revenue sharing gives a direct incentive to run bot networks that generate ad impressions and clicks.

Advertising and suspected fake traffic

  • One long-time advertiser reports:
    • Click volume suddenly rose by orders of magnitude, even after cutting bids ~30x.
    • Conversions dropped to near zero, strongly suggesting low-quality or fake traffic.
  • Others compare this to poor-converting, cheap traffic on Reddit.

Moderation, censorship, and API access

  • Commenters note:
    • Massive cuts to moderation staff and policy shifts (e.g., what speech is or isn’t disallowed).
    • Censorship/spam control is seen as unavoidable; debate is over who gets censored.
  • Locking down APIs and charging high prices for streaming access is viewed as:
    • Hindering independent bot/astroturf research.
    • Possibly a way to keep true usage/bot numbers opaque.