Analysis of FCC net neutrality comments finds over 1M pro-repeal were likely faked and 99% of organic comments were in favor of keeping rules
I used natural language processing techniques to analyze net neutrality comments submitted to the FCC from April-October 2017, and the results were disturbing.
Context & Ripple Effects
The FCC's neutrality docket had already shown signs of manipulation before this analysis: in May, its public comment API was used to submit 128K+ identical anti-net-neutrality comments, and by the August close of submissions the docket held over 21M comments, many duplicates or spambot-generated. This NLP analysis is the first attempt to separate organic signal from that noise at scale.
First-order effects
- The FCC's own rulemaking record now cuts against its proposal: if 99% of organic comments favor keeping the rules, the pro-repeal volume in the docket is largely manufactured rather than evidence of public demand for repeal.
- ISPs backing repeal lose their headline statistic — 'millions of Americans oppose net neutrality' — as an independent analysis attributes over 1M of those comments to fakery.
Second-order effects
- Opponents of repeal gain a forensic playbook: the same NLP techniques can be rerun on any docket, raising the cost of astroturfing for any firm that tries it again — a path later followed by investigations tracing fake comments to specific funders like CQ Roll Call and, ultimately, the NY AG's finding that 8.5M fake comments were funded by the largest US ISPs.
- The FCC faces pressure to authenticate its comment system, since an open API that accepted 128K identical submissions in one flood is now demonstrably a liability to whatever rule it is processing.
Third-order effects
- If bulk-faked comments stay cheap and detection stays retrospective, notice-and-comment rulemaking loses its function as a legitimacy mechanism — agencies would need identity verification at submission, and dockets become contested forensic evidence in court rather than a gauge of public sentiment.
- The pattern points toward regulatory comment fraud becoming a litigated, prosecutable issue rather than an unnoticed nuisance, with state attorneys general and academic labs doing the verification work federal agencies decline to do.
The trend: Public-comment systems are shifting from assumed-good input channels to adversarially manipulated ones, forcing agencies, courts, and investigators to treat docket volume as fabricated until verified.