CISA's new KEV list flags seven exploited flaws, two on one appliance

SonicWall's SMA1000 appliances appear twice in the seven vulnerabilities CISA added to its Known Exploited Vulnerabilities catalog. One entry covers server-side request forgery, the other OS command injection, and the two identifiers sit next to each other in sequence: CVE-2026-83548 and CVE-2026-83549.

That pairing is the most telling detail in a batch CISA says it assembled "based on evidence of active exploitation." A request-forgery flaw lets an attacker make the appliance reach somewhere it should not. A command-injection flaw on the same appliance then gives them somewhere to run, the same class of bug that put Progress LoadMaster on the list. The agency does not describe the two as a chain, and it does not have to. The list already shows the shape of one.

CVE	Product	Vulnerability class
CVE-2026-9586	Sangoma Switchvox	SQL injection
CVE-2026-48710	Kludex Starlette	HTTP request/response smuggling
CVE-2026-49869	Kestra OSS	OS command injection
CVE-2026-59822	BerriAI LiteLLM	Improper authentication
CVE-2026-82329	JFrog Artifactory	Improper authentication
CVE-2026-83548	SonicWall SMA1000 appliances	Server-side request forgery
CVE-2026-83549	SonicWall SMA1000 appliances	OS command injection

The remaining entries spread across five other products and three other vulnerability classes. BerriAI's LiteLLM and JFrog's Artifactory each appear once, both for improper authentication.

Two authentication failures in one batch is worth pausing on. Injection and smuggling flaws generally need crafted input delivered to the right component before anything happens. An authentication bypass removes that step, which is why the class tends to move quickly once it is public.

CISA's own summary of why these matter is blunt: "These types of vulnerabilities are a frequent attack vector for malicious cyber actors and pose significant risks to the federal enterprise."

What BOD 26-04 changes for federal agencies

Binding Operational Directive 26-04 governs how agencies prioritize security updates by risk, and it narrows the trigger rather than widening it. Agencies must prioritize "rapid remediation of high-risk vulnerabilities, specifically those identified by CVEs listed in CISA's KEV Catalog on publicly exposed assets that grant total control of the asset post-exploitation," while deferring action on lower-risk ones.

That is a two-part test. A KEV listing by itself does not move a flaw to the front of the queue. The affected asset has to be reachable from the internet, and exploitation has to hand an attacker the machine. Flaws outside those two conditions are treated as lower risk for sequencing.

The directive also attaches a requirement that has nothing to do with patch speed. Agencies must check whether threat actors compromised a system before the patch was applied, the same post-patch investigation the Cisco firewall entry triggered under BOD 26-04. The instruction assumes the worst case. A CVE lands in the catalog because it is being exploited now, so a fix installed today says nothing about whether someone was already inside.

What CISA requires before it adds a flaw

The catalog is not open to nominations on suspicion. CISA lists three requirements for anything submitted through its KEV nomination form: a CVE ID, evidence of exploitation, and clear mitigation guidance. The middle one does the filtering. Severe bugs stay out of the catalog regularly because nobody has shown them being used.

CISA says it will keep adding vulnerabilities that meet its criteria, and it invites outside submissions for anything it has missed. The bar there is the same. That third requirement, mitigation guidance, is easy to overlook. A flaw with no fix and no workaround is hard for an agency to action, no matter how actively it is being exploited.

BOD 26-04 binds only Federal Civilian Executive Branch agencies, and CISA says so plainly. Its ask of everyone else is softer: adopt risk-based vulnerability management and prioritize KEV entries. For a private organization, the value of the list is simpler. These seven flaws have known exploitation behind them, which is a different starting point than a severity score assigned before anyone tried.