AI Controversy Statistics
Live numbers from the SCAND.Ai monitoring pipeline, computed directly from the production database. Every statistic is dated, anchor-linked, and free to cite with attribution. How the numbers are produced: methodology. Bulk data: the AI Controversy Database.
Last computed: · License: CC BY 4.0 (cite "SCAND.Ai" with a link)
82,435 AI-industry controversy signals tracked since launch. — as of
Cite this
"82,435 AI-industry controversy signals tracked since launch." SCAND.Ai, October 25, 2026. https://scand.ai/stats#stat-topics-tracked9,778 AI controversies fully analyzed with summaries, named parties, and forecasts. — as of
Cite this
"9,778 AI controversies fully analyzed with summaries, named parties, and forecasts." SCAND.Ai, October 25, 2026. https://scand.ai/stats#stat-analyzed1,632 active AI controversies are under live monitoring right now. — as of
Cite this
"1,632 active AI controversies are under live monitoring right now." SCAND.Ai, October 25, 2026. https://scand.ai/stats#stat-active1 story is currently in the scandal state — the highest escalation level of the SCAND.Ai state machine. — as of
Cite this
"1 story is currently in the scandal state — the highest escalation level of the SCAND.Ai state machine." SCAND.Ai, October 25, 2026. https://scand.ai/stats#stat-scandals-now35 hours — the average head start between SCAND.Ai detecting a story and its first mainstream RSS coverage (measured across 105 stories). — as of
Cite this
"35 hours — the average head start between SCAND.Ai detecting a story and its first mainstream RSS coverage (measured across 105 stories)." SCAND.Ai, October 25, 2026. https://scand.ai/stats#stat-detection-speed13,664 AI public figures and organizations tracked across the industry. — as of
Cite this
"13,664 AI public figures and organizations tracked across the industry." SCAND.Ai, October 25, 2026. https://scand.ai/stats#stat-figures7,725 tracked AI figures carry full published profiles. — as of
Cite this
"7,725 tracked AI figures carry full published profiles." SCAND.Ai, October 25, 2026. https://scand.ai/stats#stat-figures-bio4,030 content items processed per day, averaged over the last 30 days. — as of
Cite this
"4,030 content items processed per day, averaged over the last 30 days." SCAND.Ai, October 25, 2026. https://scand.ai/stats#stat-daily-volume2,850 content items ingested in the last 24 hours. — as of
Cite this
"2,850 content items ingested in the last 24 hours." SCAND.Ai, October 25, 2026. https://scand.ai/stats#stat-content-24h123.1 AI-generated controversy analyses published per day, averaged over the last 30 days. — as of
Cite this
"123.1 AI-generated controversy analyses published per day, averaged over the last 30 days." SCAND.Ai, October 25, 2026. https://scand.ai/stats#stat-daily-analyses27,822 new controversy topics opened in the last 30 days. — as of
Cite this
"27,822 new controversy topics opened in the last 30 days." SCAND.Ai, October 25, 2026. https://scand.ai/stats#stat-new-topics-30d28.6% of analyzed AI controversies fall in the Ethics category — the largest single share. — as of
Cite this
"28.6% of analyzed AI controversies fall in the Ethics category — the largest single share." SCAND.Ai, October 25, 2026. https://scand.ai/stats#stat-top-category40.8/100 — the average noise level across all active AI controversies. — as of
Cite this
"40.8/100 — the average noise level across all active AI controversies." SCAND.Ai, October 25, 2026. https://scand.ai/stats#stat-avg-noise83/100 — the noise level of the loudest AI controversy right now. — as of
Cite this
"83/100 — the noise level of the loudest AI controversy right now." SCAND.Ai, October 25, 2026. https://scand.ai/stats#stat-peak-noise9,757 tracked controversies involve two or more named public figures. — as of
Cite this
"9,757 tracked controversies involve two or more named public figures." SCAND.Ai, October 25, 2026. https://scand.ai/stats#stat-multi-party1.4 distinct platforms cover the average active AI controversy. — as of
Cite this
"1.4 distinct platforms cover the average active AI controversy." SCAND.Ai, October 25, 2026. https://scand.ai/stats#stat-platforms-per-story5,947 duplicate stories were detected and merged by the six-layer deduplication pipeline in the last 30 days. — as of
Cite this
"5,947 duplicate stories were detected and merged by the six-layer deduplication pipeline in the last 30 days." SCAND.Ai, October 25, 2026. https://scand.ai/stats#stat-dedup-30d15 minutes — the fastest crawl cadence; every monitored source is re-crawled at least every 30 minutes, around the clock. — as of
Cite this
"15 minutes — the fastest crawl cadence; every monitored source is re-crawled at least every 30 minutes, around the clock." SCAND.Ai, October 25, 2026. https://scand.ai/stats#stat-crawl-cadence6 platform types are monitored continuously: Hacker News, Google News, Reddit, RSS/ArXiv feeds, Bluesky, and Polymarket prediction markets. — as of
Cite this
"6 platform types are monitored continuously: Hacker News, Google News, Reddit, RSS/ArXiv feeds, Bluesky, and Polymarket prediction markets." SCAND.Ai, October 25, 2026. https://scand.ai/stats#stat-sourcesFrequently asked questions
How many AI controversies has SCAND.Ai fully analyzed?
9,778 AI controversies fully analyzed with summaries, named parties, and forecasts.
How many AI controversies is SCAND.Ai actively monitoring right now?
1,632 active AI controversies are under live monitoring right now.
How many AI stories are currently at the scandal level?
1 story is currently in the scandal state — the highest escalation level of the SCAND.Ai state machine.
How much faster does SCAND.Ai detect AI controversies than mainstream press?
35 hours — the average head start between SCAND.Ai detecting a story and its first mainstream RSS coverage (measured across 105 stories).
How many AI public figures and organizations does SCAND.Ai track?
13,664 AI public figures and organizations tracked across the industry.
Which category has the most AI controversies?
28.6% of analyzed AI controversies fall in the Ethics category — the largest single share.
Statistics outside their plausibility bounds are automatically withheld rather than published wrong. See also: methodology · dataset · corrections policy