
Experts debate whether VulnOps can match AI-found flaws
CSA and SANS are building the model, but practitioners disagree on how much human oversight can remain.
129 stories, newest first · one card per story

CSA and SANS are building the model, but practitioners disagree on how much human oversight can remain.

Most large enterprises hold more than 1,000 untriaged findings, and about a quarter exceed 10,000.

He argues accurate supplier SBOMs speed Log4j-style response, while VEX separates vulnerable components from exploitable ones.

Timelines depend on internet access, known exploitation, whether exploitation can be automated, and attacker control.
.jpg&w=828&q=75)
Rapid7 is offering free access to the May 28, 2026 Gartner report and tying it to its own exposure-management pitch.

Cogent's AI closes vulnerability response gaps shown by PHP2Shell, where attackers exploited before patches and Fortune 500 companies patched less than 1% of 2,266 assets.

A June 2026 survey of 158 practitioners found 68.4% would be strained by 500 or more candidate findings.

Customer retests cut demonstrated compromise to zero while dozens of minor findings remained.

Qualys TRU says 357 of 48,172 flaws disclosed in 2025 were remotely exploitable and weaponized.

Brian Blakley argues attackers turn flaws into exploits faster than defenders can test, approve, and schedule patches.

High and critical findings become XSOAR incidents, and 1-Click Verify retests whether a fix is still exploitable.

Roles span penetration testing, cloud networking, ISSO compliance and SOC 2 work across eight countries.

CVE-2026-109618 and CVE-2026-109619 affect PHP, which Tenable says runs on more than 70% of websites.

Of 43 newly exploited high- and critical-severity bugs in Q3, 34 needed no authentication or user interaction.

The author says CyberGym tests found high bug-discovery accuracy but poor, often refused, automated remediation.

One CLI wraps six scan types, and a registry proxy can block known-malicious npm, Go, and Python packages before download.

Flaws without known exploitation take 107 days at the median, using February 2026 Bitsight data.

CISA has published over 67,000 CVEs and anticipates about 96,000 new vulnerabilities by year-end.

Rapid7 says multi-campus universities should share security visibility and response while keeping campus data separate.

Talos says 2026 CVE volume passed 70,000 while KEV growth lagged, so discovery is not exploitation.

Dark Reading argues vulnerability management should use environment-specific exploit evidence, not vendor scores.
Bank of America hired 1,000 contractors to patch Anthropic Mythos findings amid large AML penalties.
CISA added actively exploited Atlassian flaw CVE-2026-21589, which exposes files to external parties, to the KEV catalog.

SOCRadar released Splunk apps for MSSP incidents and vulnerability management, expanding its Splunkbase catalog to five apps.
Today 1.5K readers · 2.4K views · 754 reads · all time 189K views