MITRE Acknowledges Inaccurate Publication Timestamps in 87% of Its Recent CVE Records

MITRE acknowledges inaccurate publication timestamps in 87% of its recent CVE records and plans to retire the legacy system responsible in 2027.

  • Sarah Gooding
4 min read
MITRE Acknowledges Inaccurate Publication Timestamps in 87% of Its Recent CVE Records

MITRE has acknowledged that a legacy publishing system is responsible for inaccurate timestamps in thousands of CVE records, following research showing that 87% of the organization's CVEs published over the past year were assigned a publication time of exactly midnight UTC. MITRE says the legacy system responsible for the problem will be phased out in the first quarter of 2027.

The findings come from security researcher Jerry Gamblin, founder of RogoLabs, an independent open source research lab focused on vulnerability intelligence. His projects include the CNA Scorecard, which evaluates the completeness of CVE records, and CNAPulse, which tracks vulnerability publishing activity across CVE Numbering Authorities (CNAs). Socket has previously covered both projects and Gamblin's broader research into CVE data quality.

Analyzing CVE records published over the 52 weeks ending November 16, 2026, Gamblin found that 3,901 of MITRE's 4,509 CVEs (87%) had a datePublished timestamp of exactly 00:00:00.000Z. Across the other 425 CNAs, none of the 100,235 records carried that same timestamp.

The distribution creates the appearance that MITRE publishes hundreds of vulnerability records at precisely midnight on weekdays. In reality, those timestamps do not represent the actual time the records were published.

Publication Timestamps That Don't Match Publication Times

#

The CVE Record Format defines datePublished as the date and time a CVE record was first published in the CVE List. Researchers and automated systems can use that field to track publication activity, measure delays, and analyze vulnerability reporting trends.

Gamblin first raised the issue on GitHub in July, calling for changes to CVE publication requirements and more accurate timestamps.

To confirm the discrepancy, he compared recorded publication timestamps with the first appearance of each CVE in the official CVE List's cves/deltaLog.json change log.

For 202 MITRE records published between October 19 and November 16, he found that 144 records stamped at midnight UTC had a median delay of 15.8 hours between their recorded publication time and their first appearance in the change log. The remaining 58 MITRE records, which carried other publication times, appeared a median of 4.6 minutes after their recorded timestamps. Records from other CNAs showed a comparable median delay of 4.5 minutes.

The comparison provides evidence that the midnight timestamps are placeholders rather than actual publication times. Gamblin also found that the dates themselves were accurate in the sample he examined.

“Timestamps are one of the few fields in a CVE Record that are supposed to be objective and machine-trustworthy," Gamblin wrote in his issue on GitHub in July.

"Right now, the largest single source of CVE Records is still shipping fabricated timestamps on 84% of what it published in the last 30 days, while every other CNA in the sample gets this right at 0%. Anything downstream that measures time-to-publish, time-to-update, or disclosure latency (EPSS, CNA scorecards, research like mine) is quietly built on bad data for the biggest chunk of the dataset."

Researchers tracking CVE publication delays could interpret the 15.8-hour gap as evidence that MITRE takes much longer than other CNAs to make records available. But because the midnight timestamps are artificial, that gap cannot reliably measure publication delays. The date may still be useful for daily reporting, but the timestamp cannot reliably establish when during that day a CVE record became available.

"For me the problem is uniformity more than precision," Gamblin wrote. "What breaks things is one field that means two different things depending on who published the record."

MITRE Explains the Legacy Publishing System

#

Responding to Gamblin's findings on LinkedIn, Alec Summers, MITRE's CVE/CWE Project Leader, acknowledged the timestamp problem and attributed it to historical differences in MITRE's publishing infrastructure.

"Jerry – for clarity, this is a known historical timestamp issue that has been discussed previously, both directly and in writing," Summers wrote on LinkedIn.

Summers said MITRE publishes its CVE records through the same CVE Services API used by other CNAs, with the same validation and audit trail. However, different publishing workflows use different API endpoints.

He explained that one endpoint requires the client to supply a time value and that "because of concurrency issues in a legacy client architecture, a single predictable value (00:00:00) is best for system stability."

Summers confirmed that "Publications using that legacy architecture will be phased out in Q1 2027."

The transition is already underway. According to Gamblin's analysis, 97% of MITRE records published from 2023 through 2025 carried midnight timestamps, compared with 66% of records published so far in the fourth quarter of 2026.

Summers also addressed whether MITRE would correct historical publication timestamps.

"Modifying historical record publication timestamps, while possible, has not been highlighted as a demonstrated priority need by CVE data consumers," he wrote, adding that MITRE welcomes input on where improvements would provide the greatest value.

Gamblin, who participates in the CVE Program's Quality, Automation, and Consumer Working Groups, welcomed the Q1 2027 timeline but said he had not previously seen a documented commitment to fixing the issue. He asked MITRE to post the plan in the GitHub issue he opened in July so others following the discussion could track the progress.

CVE Publication Data Under Growing Pressure

#

In his 2026 midyear analysis, Gamblin counted 35,364 CVEs published in the first six months of the year, a 49.5% increase over the same period in 2025. Meanwhile, VulnCheck found that 23.43% of known exploited vulnerabilities identified during that period showed evidence of exploitation on or before their CVE publication date.

As disclosure volumes grow and some vulnerabilities are already being exploited when their CVE records appear, accurate publication times become increasingly useful for reconstructing vulnerability timelines. MITRE's midnight timestamps can introduce hours of error into measurements of publication latency and comparisons between CNAs, making it difficult to determine how quickly vulnerability information became publicly available.

MITRE's planned transition should improve the accuracy of future records, but thousands of historical timestamps remain inaccurate. Researchers analyzing past disclosures will still need to account for those discrepancies.

Stay ahead of threats

Subscribe to our newsletter

Get notified when we publish new security blog posts!