2026 SEC 8-K Misclassification Report — updated 2026-05-12

Body-text analysis of 4,251 SEC 8-K filings reveals that 7.3% of Item 8.01 ("Other Events") filings contain language flagging a more specific item code should have been used.

Executive summary

We analyzed 4,251 SEC Form 8-K filings filed in the most recent 21 business days and found that 7.3% of filings reported under Item 8.01 ("Other Events") contain body-text language that strongly suggests a more specific item code should have been used.

• The pattern: filers use Item 8.01 (the catch-all) when they should have used Item 1.05 (cybersecurity), 5.02 (officer departure), 1.01 (material agreement), 3.01 (notice of delisting), or 2.01 (acquisition).
• The risk: the SEC's 2023 cybersecurity disclosure rule explicitly flags Item 1.05 misclassification as a violation. Recent enforcement actions cite this exact pattern.
• The opportunity: for activist investors, journalists, compliance teams, and quant strategies — every misclassified filing is a signal that filer-trusting tools miss entirely.

Buried-event type breakdown

Of the 54 Item 8.01 filings flagged with buried events in our 21-day sample, the body language suggests:

Item code	Description	Cases
Item 1.01	Material agreement	27
Item 3.01	Notice of delisting / failure to satisfy listing rule	17
Item 1.05	Material cybersecurity incident	5
Item 5.02	Officer / director departure	3
Item 2.01	Completion of acquisition or disposition	3
Item 1.03	Bankruptcy	3

Methodology

1. Fetch every 8-K filing from EDGAR's full-index/form.idx daily feed.
2. Download the rendered HTML body of each filing, normalize to plain text via BeautifulSoup.
3. Per-item keyword + co-occurrence rules (no transformer — domain phrases are specific enough that deterministic matching beats LLM accuracy on this task). Example: Item 1.05 triggers on the co-occurrence of "material cybersecurity incident", "unauthorized access" OR "data exfiltration", AND a temporal marker ("on [date]" or "we became aware").
4. Compare body-detected items against filer-reported items in the SGML header. If the body detects an item the filer didn't report, flag as a suspected buried event.
5. Re-run nightly; results persist as a JSON column on the filing record in our parquet store.

False positive rate (manual audit of 50 flagged filings): ~12%. False negative rate is harder to measure since it requires reading every 8-K — our spot-check on 100 unflagged filings found 1 case we missed (1%).

Three specific examples (preview)

🔒 The full report has 15 named examples + the complete leaderboard. Unlock with your email →

Get the full report (free)

Enter your email to unlock the full report including:

• The complete top-50 company leaderboard with CIK + buried-event rate
• 15 named example filings with accession numbers + body excerpts
• Per-item-code methodology details + false-positive analysis
• How to monitor these in real time via the FilingFirehose API

Unlock full report

We'll send you the unlock link + a follow-up with sample API requests in 24h. No spam — unsubscribe in one click.

About the data

This analysis is built on the FilingFirehose SEC EDGAR API. We body-text-classify every 8-K, 13D, and S-3 filing in real time. The public tier (free, no auth) covers the past 72 hours; the paid tier ($29-$299/mo or $999 lifetime) covers the full historical archive with webhook alerts.

• Original research piece
• Live company leaderboard
• Today's flagged filings
• Public REST API (no auth)