Prolepsia forecastYes / no

Will CISA add more than 200 vulnerabilities to its Known Exploited Vulnerabilities catalog in 2027?

As of 17 November 2026, Prolepsia puts the chance of yes at 95%.

95%chance of yes

Yes 95%No 5%
Prolepsia forecast made

How Prolepsia reasons

The report Prolepsia wrote with this forecast.

How forecasts are made · Track record

TL;DR

The forecast is 94.8% Yes and 5.2% No that CISA adds at least 201 distinct vulnerabilities to its KEV catalog during 2027. Recent cataloging activity provides a substantial margin above the threshold; a sustained decline in throughput is the main route to a No outcome.

Key evidence

  • Recent throughput is well above the threshold. The official November 13, 2026 snapshot records 187 additions dated in 2023, 186 in 2024, 245 in 2025, and 297 during January 1–November 13, 2026. The earlier below-threshold years matter, but the recent direction is clear (official catalog).
  • The increase extends beyond a single batch. January–October 2026 contained 282 additions, an annualized pace of 338.4. Finishing a year at only 200 would represent approximately a 41% reduction from that pace. Growth also remained after excluding Microsoft and each year's largest addition day (catalog data).
  • CISA is maintaining the intake pipeline. Its May 21, 2026 announcement introduced a secure nomination form while retaining the requirements for a CVE identifier, exploitation evidence, and mitigation guidance. This supports continuity, but does not guarantee throughput (CISA announcement).
  • KEV retains an operational role. CISA's risk-based remediation directive continued using KEV status, while the notice ending a separate weekly vulnerability bulletin explicitly directed readers toward KEV (directive announcement; bulletin notice).

Uncertainties

The key uncertainty is whether the recent cataloging pace persists. Two completed recent years fell below the threshold, and the mature historical record is short. Funding proposals and agency-wide workforce figures do not establish the resources available to the KEV team.

The outcome measures gross additions, not net catalog growth or newly discovered vulnerabilities. Current-snapshot historical counts omit entries subsequently removed; final resolution must account for verified qualifying removals and corrected official addition dates. These counting details matter, but sustained operational or policy deterioration is the larger risk.

Read the full report

TL;DR

The forecast is 94.8% Yes and 5.2% No that CISA adds at least 201 distinct vulnerabilities to its KEV catalog during 2027. Recent additions run well above the threshold, and the increase extends across months and vendors. Treat Yes as the clear expected outcome and sustained loss of cataloging capacity as the main threat to it.

Context

This question measures entries added to KEV, not vulnerabilities discovered during the year. Each distinct CVE first added between January 1 and December 31, 2027 counts once. CISA's schema defines dateAdded as the date of catalog inclusion, rather than discovery or remediation timing (official schema). Under the resolution rules, metadata changes do not count, verified additions later removed still count, and corrected official dates govern eligibility.

As of the November 17, 2026 forecast date, the latest verified catalog release was November 13, 2026, containing 1,781 records. The observed addition history therefore ends on November 13, not on the forecast date (official catalog). Resolution ordinarily occurs after January 31, 2028, allowing time for late updates and corrections under the question's rules.

Evidence

The historical record gives two different lessons. KEV has supported annual additions far above the threshold, but below-threshold years are also recent. The full annual history available in the current snapshot is shown below. These figures count entries retained in that snapshot, grouped by their official addition dates; they are not complete reconstructions of historical gross additions (official catalog).

Addition period Entries retained in the snapshot
November–December 2021, partial launch year 311
2022 555
2023 187
2024 186
2025 245
January 1–November 13, 2026 297

The launch and early backfill period is a weak guide to normal operations. The more relevant completed-year comparison is 2023–2025: only one of those three years exceeded 200. That is the strongest historical caution against treating Yes as automatic. But giving those years equal weight would miss the subsequent change in throughput. The observed 2026 total had already passed the threshold well before year-end (catalog history).

The complete monthly history supplied for the mature period makes that change visible. November 2026 is partial through November 13; December 2026 is not observed. All figures below come from the same catalog snapshot (official addition-date data).

Month 2023 2024 2025 2026
January 5 21 14 17
February 14 9 27 28
March 18 10 32 26
April 17 10 15 31
May 19 14 24 21
June 24 9 20 23
July 16 14 20 26
August 8 19 15 31
September 19 25 16 43
October 18 17 31 36
November 18 22 11 15, partial
December 11 16 20 Not observed

January–October 2026 contained 282 additions, equivalent to 338.4 over twelve months. This is a pace comparison, not a prediction of the eventual count. Reaching only 200 over a full year would require approximately a 41% reduction from that pace. The table also shows that elevated activity preceded the September–October peak: the case does not depend on assuming those two months repeat (catalog data).

The composition of the increase strengthens the case for persistence. The research's catalog diagnostics found that growth remained after excluding Microsoft and each year's largest addition day. They also found broader vendor coverage and a lower share of entries carrying earlier-year CVE identifiers. I read this as evidence against a single-vendor spike or a one-off sweep of old records. CVE identifier year remains only an age proxy; it does not establish when a flaw was disclosed or first exploited (underlying catalog).

There is also a concrete intake improvement. On May 21, 2026, CISA announced a secure KEV nomination form intended to improve intake and analysis. The notice preserved the requirements for a CVE identifier, credible exploitation evidence, and mitigation guidance. This supports the existence of a continuing pipeline. It does not show a measured capacity increase or prove that the form caused the higher addition rate (nomination-form announcement).

Policy signals point toward continued use of KEV. CISA's June 10, 2026 directive retained KEV status in federal remediation prioritization. Its September notice ending the broad weekly vulnerability bulletin directed subscribers toward KEV rather than announcing KEV's withdrawal. FedRAMP's implementation guidance adds another operational use for exploitation-based prioritization. These are signs of demand for a maintained catalog, not promises of a minimum annual count (directive; bulletin notice; FedRAMP guidance).

The institutional evidence is mixed. Agency-wide staffing and funding pressures are real, but headline cuts cannot simply be assigned to KEV. The House appropriations report proposed higher funding for Vulnerability Management than its prior-year comparator even while recommending reductions elsewhere. A proposal is not enacted funding, and a broad program allocation is not KEV staffing (House report). Shutdown evidence also cuts both ways: agency operations were impaired, yet the monthly catalog history records continued additions during the disruption (congressional testimony; catalog).

New reporting channels offer a supporting mechanism, not a decisive reason for Yes. European mandatory manufacturer reporting began on September 11, 2026. That can expand the pool of exploitation evidence, but confidential submissions, disclosure delays, and different eligibility rules separate a European report from an accepted CISA entry. No measured flow from that system into KEV was established (ENISA announcement). Likewise, faster CVE ingestion or NVD enrichment is not the same as verifying exploitation and admitting a vulnerability to KEV (NIST update).

What's non-obvious

The useful signal is not simply that vulnerability activity is rising. This outcome depends on a narrower process: exploitation evidence must reach CISA, meet its criteria, and become a catalog entry. The stronger evidence is that observed additions increased across months and remained higher after removing obvious concentration effects. That bears more directly on this question than general growth in CVE submissions or cybersecurity spending (catalog; NIST context).

A second distinction is between catalog size and gross additions. A current file can undercount the latter because entries can be removed. The historical archive review identified such cases, but did not fully audit every addition-date correction. The right conclusion is a measurement caution, not a reason to treat the recent pace as weaker. Final counting must use official history where necessary, rather than subtracting the starting catalog size from the ending size (official repository).

Uncertainties

The largest uncertainty is rate persistence. The recent pace provides room for a decline, but the earlier below-threshold years show that lower throughput is possible. The mature record is too short to pin down how often a sustained reversal occurs. Continued exploitation alone does not guarantee unchanged cataloging activity.

Dedicated KEV staffing, nomination backlog, acceptance rates, processing delays, and a KEV-specific shutdown operating plan were not established. Those data would help distinguish durable capacity from temporary high output. Final funding and actual staffing execution would also be more informative than proposed agency totals.

A narrower inclusion policy, prolonged staffing loss, impaired evidence sharing, or discontinuation before enough entries accumulate can produce a No outcome. Discontinuation itself does not annul the question; qualifying additions made beforehand still count under its rules. These risks explain the downside probability without changing the clear direction of the forecast.

Forecast history

How Prolepsia’s forecast has movedOne forecast so far: 95% on 17 Nov 2026.
DateForecast
95% chance of yes

Sources Prolepsia read

  1. official schemaraw.githubusercontent.com
  2. ​​CISA Releases New Known Exploited Vulnerability Nomination Formcontent.govdelivery.com
  3. FedRAMP Response to CISA BOD 26-04 (Prioritizing Security Updates Based on Risk)www.fedramp.gov
  4. 78 Mission The Cybersecurity and Infrastructurewww.govinfo.gov
  5. congressional testimonyhomeland.house.gov
  6. The CRA Single Reporting Platform is launched | ENISAwww.enisa.europa.eu
  7. NIST Updates NVD Operations to Address Record CVE Growth | NISTwww.nist.gov
  8. GitHub - cisagov/kev-data: Mirror of cisa.gov/kev data files · GitHubgithub.com
  9. Vulnerability Report - September 2026 – Vulnerability-Lookupwww.vulnerability-lookup.org
  10. Grassley to DHS - Shutdownwww.grassley.senate.gov

The question

This question asks whether the U.S. Cybersecurity and Infrastructure Security Agency (CISA) will add at least 201 distinct vulnerabilities to its Known Exploited Vulnerabilities (KEV) catalog during the 2027 calendar year, from January 1, 2027, through December 31, 2027, inclusive. The KEV catalog is maintained by CISA and identifies vulnerabilities for which there is credible evidence of exploitation in the wild. CISA publishes the catalog in machine-readable JSON and CSV formats, with each vulnerability identified by a Common Vulnerabilities and Exposures (CVE) identifier and assigned a dateAdded field indicating when it was included in the catalog. ([github.com](https://github.com/cisagov/kev-data)) As of November 17, 2026, historical data indicates that CISA added approximately 186 vulnerabilities in 2024 and 245 in 2025. During 2026, the rate of additions has accelerated, with approximately 297 vulnerabilities added through November 13, 2026. These figures suggest that exceeding 200 additions in a calendar year is plausible, although annual totals vary with exploitation activity, vulnerability disclosures, and CISA's cataloging practices. ([invadel.com](https://invadel.com/blog/vulnerability-statistics/)) The question concerns the number of vulnerabilities newly added to the catalog during 2027, regardless of when those vulnerabilities were originally discovered, assigned CVE identifiers, or first exploited. It does not concern the total number of vulnerabilities present in the catalog at the end of 2027 or the number of vulnerabilities first discovered in 2027.

How it resolves

The question resolves YES if CISA adds 201 or more distinct vulnerabilities to its official Known Exploited Vulnerabilities catalog with addition dates between January 1, 2027, and December 31, 2027, inclusive. The question resolves NO if CISA adds 200 or fewer distinct vulnerabilities during that period. The primary resolution source will be CISA's official Known Exploited Vulnerabilities catalog, available through the CISA website and its official GitHub repository (https://github.com/cisagov/kev-data). CISA's machine-readable JSON or CSV dataset will be used to calculate the number of qualifying entries. Each distinct CVE identifier with a dateAdded value within the 2027 calendar year will count as one addition. The official dataset defines dateAdded as the date on which the vulnerability was added to the catalog. ([github.com](https://github.com/cisagov/kev-data)) Resolution will ordinarily occur after January 31, 2028, using the latest available official CISA catalog at that time. This provides time for CISA to publish late updates or corrections to entries added during December 2027. If the official catalog remains available, the number of qualifying entries will be determined by filtering for dateAdded values from 2027-01-01 through 2027-12-31, inclusive, and counting unique CVE identifiers. If CISA changes the structure or format of the catalog, the count will be determined using equivalent official fields or CISA's published historical records. If the catalog is discontinued during 2027, qualifying additions made before discontinuation will still count, and discontinuation itself will not trigger annulment. If the catalog becomes temporarily unavailable, CISA's official GitHub repository and historical versions will be used to reconstruct the additions. The question will be annulled only if reliable official records are insufficient to determine whether the threshold of 201 additions was reached, including after consulting available official archived versions and publication records.

Fine print

Each distinct CVE identifier will count once, even if CISA updates its entry multiple times during 2027. Changes to descriptions, severity information, remediation instructions, deadlines, or other metadata do not count as additional vulnerabilities. Vulnerabilities originally disclosed or assigned CVE identifiers before 2027 will count if they are first added to the KEV catalog during 2027. Conversely, vulnerabilities assigned CVE identifiers in 2027 but not added to the catalog until 2028 or later will not count. If a vulnerability was added before 2027 and subsequently modified or reintroduced during 2027, it will not count as a new addition unless CISA explicitly identifies it as a distinct vulnerability with a new CVE identifier first added in 2027. If CISA removes a vulnerability that was originally added during 2027, it will still count toward the annual total, provided its original addition can be verified through official historical records. The question measures gross additions rather than the net increase in catalog size. If CISA retroactively corrects a dateAdded value, the corrected official value available at the resolution cutoff will determine whether the vulnerability counts. If multiple official sources disagree, CISA's current official dataset takes precedence over its GitHub mirror, followed by historical official records where necessary. The threshold is strictly greater than 200. Exactly 200 additions resolves NO, while 201 additions resolves YES. Calendar dates are interpreted using the dates officially assigned by CISA in the catalog rather than timestamps converted into a particular time zone.