The Race to Remediation: Black October Showed that Threat Intelligence Isn’t Enough

In a single week in October 2026, security teams faced one serious security incident after another. A certificate authority failed, critical flaws in PHP were disclosed after a broken embargo, a popular open-source AI gateway was compromised, and developer platforms reported thefts of customer secrets. Industry insiders started calling the seemingly steady drumbeat of attacks “Black October.”

Any one of these incidents would have kept a security team busy. Collectively, they exposed a bottleneck that has little to do with any single exploit. Security teams are not short on vulnerability information. They are short on answers: what is affected in our environment, and how can we fix it – fast?

As JFrog CEO Shlomi Ben Haim put it: “A single weak link can cost far more than the fix. For LiteLLM and PHP2Shell, being caught off-guard means money, downtime, and customer trust. The companies that come through well are the ones that know exactly what they are running and can act within hours vs. days. JFrog’s role is to find the problems early, tell customers what to do, and make sure they can precisely determine their exposure.”

Three Pressures That Compound

Three forces explain why answers are getting harder to produce:

  • Volume: In the AI-accelerated landscape of 2026, based on our count of NIST National Vulnerability Database (NVD) records, 36,952 CVEs were published in Q3 2026 alone, more than triple the 11,725 published in Q3 2025 (as of November 16, 2026).
  • Exploitability: A severity score does not say whether a flaw is applicable in your environment. Triage by count alone wastes scarce resources and time.
  • Attacker Speed: A Cloud Security Alliance survey of more than 900 security leaders found that only 9% of organizations remediate critical or high-severity production vulnerabilities within 24 hours, while 74% take between one and seven days. Meanwhile, attackers can use AI to speed up exploitation, chaining together low-severity findings and making repeated automated attempts until one unpatched system gives them an opening.

That is a different kind of weakness, and it needs a different kind of approach. A faster version of the old process will not close it. AI pushes on all three pressures: it surfaces more flaws and drafts more fixes, but maintainers cannot absorb unlimited volume.

The Disclosure Process Wasn’t Built for This

The vulnerability disclosure process was designed around human review cycles. A researcher reports a flaw, a maintainer triages it, a fix is written, and a CVE is published. That rhythm worked when exploitation was also a manual effort. It does not hold up when attackers can automate discovery and exploitation at scale.

Enterprises have also historically depended on upstream maintainers to publish patches on their own timelines, and those timelines are asynchronous with the CVE disclosure window. In an AI-powered threat landscape, that mismatch becomes an exposure gap that attackers can use.

A new ecosystem of patching vendors has stepped in to help close it. These vendors produce signed, backported fixes for the exact versions enterprises actually consume, often ahead of public disclosure.

Why Does Threat Intelligence Alone Leave a Gap?

Threat intelligence describes the outside world, not your software environment. That difference explains why teams with excellent security vulnerability feeds still lose hours after a disclosure. Two gaps open between a published threat and a confident response:

Timing: Public databases trail events. As of October 2026, the National Institute of Standards and Technology (NIST) now enriches only a subset of CVEs, and PHP2Shell was being probed days before patches shipped. A process that waits for them is structurally late.

Remediation: Even with timely, contextual intelligence, someone still has to find a trustworthy fix, apply it across every affected artifact (assuming all artifacts are even traceable), and prove it worked. This is where many teams slow from hours to days.

Know Your Exposure With JFrog Threat Hub

Finding problems early and determining exposure precisely is the first half of the job. JFrog Threat Hub, a feature of JFrog Xray, brings JFrog Security Research findings and vetted external advisories into the same single source of truth that holds your scanned artifacts, so a published threat can be matched to your own environment in one step.

Knowing your exposure only matters if you can act on it quickly, which is the second half of the job.

Fortified Defenses Start With One Source of Truth

A month like October does not call for a faster version of the old process. It calls for a different one. In our view, LiteLLM and PHP2Shell showed what separates the teams that come through well: they could see what they run, and they could fix it fast.

Know everything you run

Start with an inventory at every layer: source code, packages, container images, AI models, and agent tooling. A component you cannot see cannot be prioritized or patched. Then verify what you deploy, not only what you wrote, because malicious code can live in built artifacts while the source stays clean. JFrog Artifactory serves as that single source of truth, holding the main artifact types (binaries, containers, AI models, MCP server packages) in one system of record.

Draw on an ecosystem of fix providers

Fix providers often specialize in particular software ecosystems, so drawing on an ecosystem of providers can help broaden coverage. In our view, reducing this risk requires two critical capabilities: high-quality patches across software ecosystems and a control plane designed to automate remediation at the binary level.

Close the loop with zero-touch remediation

A patch sitting in a vendor’s repository does nothing for an artifact already running in production. Artifactory can act as the control plane that ingests each fix from supported partners. JFrog Security can then apply a selected fix according to the customer’s policy and context, and produce a signed evidence chain that can support audit reviews. The goal is a self-healing software supply chain and zero-touch remediation that aims to help customers patch in hours rather than days. For branches past their end of life, migration remains the durable fix.

Ready for the Next Black October?

When the next disclosure lands, what is affected, and can we fix it in hours rather than days? Organizations that can provide a quality answer will weather the next crowded month. Those that cannot will spend it searching.

Move From Disclosure to Fix in Hours

Patching at scale takes a single source of truth, trusted fixes from an ecosystem of partners, and a universal control plane that applies them automatically. See how zero-touch remediation and the self-healing software supply chain work in your own environment: schedule a demo, take an online tour, or start a free trial of the JFrog Platform.