What Is a Zero-Day Vulnerability? A Complete Guide
What a zero-day vulnerability is, how attackers find and exploit one, and how patch management, EDR, and the CISA KEV catalog help you defend against it.

A zero-day vulnerability is a software or hardware flaw that is exploited by attackers before the vendor knows about it or has released a fix — meaning defenders have had “zero days” to prepare. Because no patch exists yet, every device or system running the affected software is exposed, no matter how diligent the owner is about updates. Zero-days are the most prized weapons in an attacker's arsenal, used by nation-state hacking groups, ransomware crews and spyware vendors alike, and they sit at the center of the modern vulnerability-management conversation that includes patch cycles, endpoint detection and response (EDR) tools, and the government watchlists that tell IT teams what to fix first.
Quick facts: zero-days at a glance
- A zero-day has no vendor patch available at the time it is first exploited.
- Google's Threat Intelligence Group tracked 90 zero-days exploited in the wild in 2025, versus 75 in 2024.
- The U.S. Cybersecurity and Infrastructure Security Agency's Known Exploited Vulnerabilities (KEV) catalog listed 1,786 confirmed-exploited flaws as of November 2026.
- Most real-world breaches still come from known, unpatched vulnerabilities, not zero-days — patching on time remains the single biggest defense.
What is a zero-day vulnerability?
A zero-day vulnerability is a security weakness in software, firmware or hardware that becomes known to attackers — or is actively being exploited — before the company responsible for the product has a fix ready. The “day zero” is the moment the flaw is first identified by someone outside the vendor's own testing process, whether that's a criminal group, a government intelligence agency or an independent researcher. Everything that happens afterward — reporting, patching, disclosure — is counted in days from that starting point.
The term covers two closely related ideas: the underlying zero-day vulnerability (the flaw itself) and the zero-day exploit (the code or technique an attacker uses to take advantage of it). A zero-day attack is any intrusion that relies on that exploit before a patch exists. Once the vendor ships a fix, the vulnerability stops being a zero-day — it becomes a known, or “n-day,” vulnerability, even if many systems remain unpatched and still exploitable for months or years afterward.
The zero-day lifecycle: discovery to patch
Security researchers generally describe a zero-day's life in four stages:
- Discovery. Someone — a bug-bounty researcher, an internal security engineer, a nation-state analyst, or a criminal group — finds the flaw. This is “Day 0.”
- Reporting (or weaponization). A researcher who intends to help typically reports the bug privately to the vendor or a coordinating third party, such as a CERT. An attacker who finds it first instead builds an exploit and uses it quietly, often for months, before anyone else notices.
- Patch development. Once the vendor is aware, engineers build, test and release a fix. This can take anywhere from a few days for a simple bug to several weeks or months for a flaw buried deep in core infrastructure software.
- Public disclosure. Details are published once a patch is available, so defenders can act on specific, verified information rather than rumor. Many researchers follow a standard disclosure deadline — commonly 90 days from the initial report — and will publish their findings even without a vendor patch if that window lapses, to pressure fixes and keep the public informed.
The gap between stage one and stage three is the most dangerous window: the vulnerability is known to at least one party, but no official fix exists. A 2017 RAND Corporation study of a nonpublic zero-day stockpile found that it took an average of 22 days to develop a working exploit once a vulnerability was found, and that zero-day exploits had an average “shelf life” of 6.9 years before public disclosure or patching — a reminder that many zero-days are discovered and then held, not disclosed, for a long time.
How are zero-days discovered?
Zero-days surface through several very different channels, each with different incentives:
- Bug bounty and vulnerability research programs. Independent researchers and security firms hunt for flaws in popular software and report them to the vendor, often in exchange for a bounty, through formal vulnerability disclosure programs.
- Vendor and internal security teams. Companies run their own fuzzing, code audits and red-team exercises, catching some flaws before anyone outside ever sees them.
- Commercial exploit brokers. Firms that specialize in finding and selling zero-days to governments and other paying customers operate a legal, if controversial, market separate from criminal exploit sales.
- Government and intelligence agencies. National security services actively search for and stockpile zero-days for offensive and defensive use. In the United States, the Vulnerabilities Equities Process (VEP) — formalized in a November 2017 White House charter — governs whether the government discloses a vulnerability to the affected vendor or withholds it for intelligence or law-enforcement purposes.
- Incident response and threat intelligence. Sometimes a zero-day is only discovered after it has already been used in an attack, when forensic investigators or security vendors analyze malware and traffic patterns and find an exploit for a flaw nobody had reported yet.
Attackers who find a flaw first have every incentive to keep it secret for as long as possible — the longer a zero-day stays unknown to the vendor, the longer it keeps working.
How are zero-days exploited?
Once an attacker has a working exploit, it is delivered the same way any other attack is: through phishing emails and malicious attachments, booby-trapped websites, compromised software updates (a supply-chain attack), malicious documents, or direct exploitation of an internet-facing server. What makes a zero-day distinct is simply that no signature, patch or vendor advisory exists yet to stop it — traditional defenses built around known indicators of compromise are far less effective.
Zero-days are disproportionately used in targeted, high-value operations rather than mass, indiscriminate attacks, because they are expensive to find or buy and lose their value the moment they are burned (detected and patched). That's why most advanced persistent threat (APT) campaigns against governments, defense contractors, journalists and dissidents lean on zero-days, while the much larger volume of everyday cybercrime — ransomware, credential theft, business email compromise — overwhelmingly exploits vulnerabilities that already have patches available but simply haven't been applied.
Zero-day vs. a known (already-patched) vulnerability
The practical difference comes down to whether a fix exists and whether it's been applied:
| Characteristic | Zero-day vulnerability | Known (n-day) vulnerability |
|---|---|---|
| Vendor patch available? | No — not yet released | Yes, but may not be installed everywhere |
| Who knows about it | Only the discoverer (and anyone they've told) | Publicly documented, usually with a CVE ID |
| Primary users | Nation-states, APT groups, exploit brokers, high-end spyware vendors | Broad range, including opportunistic criminal groups and botnets |
| Detection difficulty | High — no signatures or indicators exist yet | Lower — security tools can be updated with known indicators |
| Best defense | Behavior-based detection (EDR/XDR), network segmentation, least privilege | Timely patching, vulnerability scanning, KEV-driven prioritization |
| Share of real-world breaches | A minority, though a high-impact one | The large majority of breaches trace back to unpatched known flaws |
This is also why security teams distinguish between a vulnerability being “zero-day” and being “critical.” A known, unpatched vulnerability — even one that's been public for years — can be just as damaging as a fresh zero-day if an organization never applies the fix. That's the core logic behind CISA's KEV catalog, discussed below: it exists precisely because attackers keep successfully reusing old, already-fixed bugs against organizations that never patched them.
Responsible (coordinated) disclosure, explained
“Responsible disclosure,” more formally called coordinated vulnerability disclosure, is the practice of privately reporting a flaw to the vendor and giving them a reasonable window to fix it before any public details are released. The goal is to balance two competing interests: the public's right to know about risks to their systems, and the need to avoid handing attackers a ready-made blueprint before a fix ships.
In practice, coordinated disclosure typically looks like this: a researcher privately reports the bug, the vendor confirms and assigns a severity rating, a patch is developed and tested, and only then — often alongside the patch release — is a public write-up and a CVE identifier published. If a vendor goes silent or stalls well past an agreed deadline, many researchers will disclose anyway, arguing that public pressure and defender awareness outweigh the risk of tipping off attackers. Government disclosure follows its own process: in the U.S., the Vulnerabilities Equities Process weighs whether a vulnerability found by a federal agency should be reported to the vendor or kept secret for intelligence purposes.
Real-world zero-day examples
A few documented cases illustrate how differently zero-days can be used:
- Stuxnet (discovered 2010). This worm used four separate Windows zero-day vulnerabilities to reach and sabotage programmable logic controllers running Siemens industrial software at Iran's uranium-enrichment facilities. Security firm Symantec traced development activity back to at least 2007, and found the worm had spread to more than 100,000 computers worldwide by September 2010. No government has officially claimed responsibility, though it is widely reported to have been a joint U.S.-Israeli operation.
- Log4Shell / CVE-2021-44228 (disclosed December 2021). A critical flaw in the widely used Apache Log4j 2 Java logging library let attackers run arbitrary code simply by getting a malicious string logged — for example, in an HTTP header. It carried the maximum possible CVSS severity score of 10.0 and affected an enormous range of enterprise software and cloud services, prompting CISA's director at the time to call it one of the most serious vulnerabilities of her career.
- EternalBlue and the Shadow Brokers leak (2016–2017). A group calling itself the Shadow Brokers released a cache of NSA-linked exploit tools, including EternalBlue, which targeted a Windows SMB flaw. It was later repurposed in the WannaCry and NotPetya attacks, which caused billions of dollars in global damage.
- Microsoft Exchange Server / ProxyLogon (2021). A Chinese state-sponsored group tracked as Hafnium exploited a chain of zero-days in on-premises Microsoft Exchange Server, compromising tens of thousands of organizations' mail servers before patches and detection tools caught up.
These are the dramatic, headline-grabbing cases — but the much larger volume of zero-day and KEV-listed activity today is narrower in scope: individual flaws in enterprise software, like a remote-code-execution bug recently added to the KEV catalog in a popular AI tooling library, that still demand the same urgency from IT teams even without making front-page news.
The CISA KEV catalog: your patch priority list
With thousands of new CVEs published every year, most organizations can't patch everything immediately — so the U.S. Cybersecurity and Infrastructure Security Agency maintains the Known Exploited Vulnerabilities (KEV) catalog, which it describes as “the authoritative source of vulnerabilities that have been exploited in the wild.” Each entry lists the CVE ID, the date it was added, a remediation due date, and whether it's known to be used in ransomware campaigns. As of November 2026 the catalog listed 1,786 entries, available as a browsable list and as a machine-readable JSON feed that security teams can pull directly into their vulnerability management tools.
Under CISA's binding operational directive framework, federal civilian agencies are required to remediate KEV-listed flaws by their assigned due dates, and agencies are responsible for evaluating how exposed their own internet-facing assets are. CISA explicitly recommends that every organization, not just federal agencies, use the KEV catalog “as an input to their vulnerability management prioritization framework” — in other words, when you can't patch everything at once, patch what's on the KEV list first, since those are flaws attackers are demonstrably already using. A recent example close to home: a November 2026 Patch Tuesday release that fixed an actively exploited Outlook zero-day is exactly the kind of flaw that tends to land on the KEV list soon after disclosure.
How individuals can protect themselves
- Turn on automatic updates. Operating systems, browsers and mobile apps patch most zero-days within days of disclosure — automatic updates close that window as fast as possible.
- Don't delay OS and app updates. The gap between a patch being available and being installed is where most real-world damage happens, zero-day or not.
- Use a password manager and multi-factor authentication. Many zero-day attack chains still need a second step, like credential theft, to reach their real target — strong authentication breaks that chain.
- Be skeptical of unsolicited attachments and links. Phishing remains the most common delivery mechanism for both zero-day and known exploits alike.
- Know what to do if you're compromised. If you suspect your accounts or data were exposed, follow a clear incident response checklist rather than guessing.
How organizations can protect themselves: patch management and EDR
No single control stops zero-days — defense works in layers:
- Disciplined patch management. A formal process to track, test and deploy vendor patches quickly, prioritized using the CISA KEV catalog and internal asset criticality, closes the known-vulnerability gap that causes the majority of breaches and shrinks the window a zero-day stays unpatched once a fix ships.
- Endpoint Detection and Response (EDR) and XDR. Because zero-days have no existing signature, behavior-based tools that watch for suspicious process activity, privilege escalation and unusual network connections can catch an exploit in action even when nothing yet recognizes the specific vulnerability.
- Network segmentation and least privilege. Limiting what a compromised system can reach, and what a compromised account can do, contains the blast radius of a successful zero-day exploit.
- Virtual patching and compensating controls. Web application firewalls, intrusion prevention systems and vendor-published mitigations can reduce exposure in the gap between disclosure and an installable patch.
- Threat intelligence feeds and the KEV catalog. Subscribing to vendor security advisories and the CISA KEV feed gives security teams an early, prioritized signal about which vulnerabilities are actively being exploited right now, rather than relying on severity scores alone.
- A vulnerability disclosure program. Giving external researchers a clear, safe channel to report flaws — ideally with a bug bounty — increases the odds your own vulnerabilities are found by a helpful researcher before a criminal group finds them first.
Bottom line
A zero-day vulnerability is dangerous specifically because no fix exists yet — but the uncomfortable truth, backed by years of breach data, is that most real-world damage still comes from known vulnerabilities that already have a patch sitting unapplied. The practical response to both problems looks the same: patch fast, prioritize using authoritative sources like the CISA KEV catalog, deploy behavior-based detection that doesn't depend on recognizing a specific exploit, and support coordinated disclosure so researchers have an incentive to report flaws to vendors instead of selling or hoarding them. Zero-days will keep making headlines — Stuxnet and Log4Shell both made the case loudly — but disciplined basics close more of the real risk than chasing the next exotic exploit ever will.
Frequently asked questions
What is a zero-day vulnerability in simple terms?
It's a security flaw that attackers can exploit before the software maker knows about it or has released a fix, so there is no patch available to protect against it yet.
What's the difference between a zero-day vulnerability and a zero-day exploit?
The vulnerability is the underlying flaw in the code. The exploit is the specific tool or technique attackers write to actually take advantage of that flaw. A zero-day attack is an intrusion that uses a zero-day exploit.
How is a zero-day different from a regular, already-patched vulnerability?
A regular vulnerability has a vendor patch available, even if some systems haven't installed it yet. A zero-day has no patch at all at the time it's first exploited, so every system running that software is exposed until a fix ships.
What is the CISA KEV catalog and why does it matter?
The Known Exploited Vulnerabilities (KEV) catalog is a list maintained by the U.S. Cybersecurity and Infrastructure Security Agency of vulnerabilities confirmed to be actively exploited in the wild. CISA recommends every organization use it to prioritize which flaws to patch first, since those are the ones attackers are demonstrably already using.
Can antivirus or EDR software stop a zero-day attack?
Traditional antivirus that relies on known signatures usually can't catch a brand-new zero-day exploit. Endpoint Detection and Response (EDR) tools improve the odds because they watch for suspicious behavior, like unusual process activity or privilege escalation, rather than only matching known threats.
What is responsible disclosure of a vulnerability?
Responsible, or coordinated, disclosure means a researcher privately reports a flaw to the vendor and gives them time to fix it before publishing details, so attackers don't get a ready-made blueprint before a patch exists.
Sources
- Wikipedia: Zero-day vulnerabilityen.wikipedia.org
- Wikipedia: Stuxneten.wikipedia.org
- Wikipedia: Log4Shellen.wikipedia.org
- CVE.org record: CVE-2021-44228 (Log4Shell)cve.org
- CISA: Known Exploited Vulnerabilities Catalogcisa.gov
Sana Qureshi runs the security and privacy desk. She reports on actively exploited vulnerabilities, vendor patches and data breaches, and covers the password managers, VPNs and authentication tools readers use to protect themselves. Her alerts cite vendor advisories, CISA and the CVE record directly.

