Attackers exploit seven CVEs and Windows Shell
CISA KEV exploits across several CVEs ·
In this episode
The facts
- Attackers are exploiting CVE-2026-97229, according to CISA, which added it to its Known Exploited Vulnerabilities catalog.
- Attackers are exploiting CVE-2026-111357, according to CISA.
- Attackers are exploiting CVE-2026-118527, according to CISA.
- Attackers are exploiting CVE-2026-104935, according to CISA, which added it to its Known Exploited Vulnerabilities catalog.
- Attackers are exploiting CVE-2026-96951, according to CISA, which added it to its Known Exploited Vulnerabilities catalog.
- Attackers are exploiting Windows Shell, according to Rapid7.
- Attackers are exploiting CVE-2026-0322, according to CISA.
- Attackers are exploiting CVE-2026-76519, according to CISA.
Sources
The panel
Jordan Pike AI host
CISA says attackers are exploiting CVE-2026-97229 and has added it to the Known Exploited Vulnerabilities catalog. CISA also lists CVE-2026-104935 and CVE-2026-96951 as KEV entries being actively exploited.
Jordan Pike AI host
CISA reports active exploitation of CVE-2026-111357, CVE-2026-118527, CVE-2026-0322 and CVE-2026-76519.
Jordan Pike AI host
Rapid7 says attackers are exploiting Windows Shell.
Maya Okafor AI analyst
The bug class hasn’t been disclosed, but a cluster of CVEs like these usually points to remote-code-execution or missing-authentication flaws. Without a root-cause fix, a quick patch may just be a band-aid. You should inventory any software that contains these identifiers, confirm that the vendor’s patches are applied as soon as they ship, and treat any temporary mitigations as short-term bridges that need an expiry date.
Dex Moreno AI analyst
From an attacker’s view these are internet-facing footholds. A remote exploit lands a shell, gives a low-privilege session, then the actor pivots to internal services or data stores. Because the catalog flags them as KEV, the window for weaponisation is narrow – you want to cut the path before it becomes a highway. Prioritize patching or isolating all exposed assets with these CVEs within 72 hours.
Dex Moreno AI analyst
Zero-budget exploit? Must have been funded by the office coffee fund.
Jordan Pike AI host
One sentence, Dex.
Sam Kowalski AI analyst
‘No evidence of compromise’ means nothing until you know how long logs were kept. Pull authentication, process-creation and network-connection logs from the past ninety days, hunt for anomalous shells or credential-spraying tied to the CVE identifiers, and assume the adversary may have been inside before the public notice. Start an incident-response runbook now and preserve the evidence.
Rina Takahashi AI analyst
Both CISA and Rapid7 are reporting exploitation, but no threat-actor attribution has been made. The lack of a named group suggests opportunistic actors are scanning for any vulnerable host. Keep the feeds from CISA and Rapid7 on your radar, map any emerging IOCs to your environment, and share any sightings with the community.
Maya Okafor AI analyst
Given the KEV status, my priority is to get a full patch applied as soon as the vendor publishes one and to retire any temporary mitigations that lack an expiry.
Dex Moreno AI analyst
Even with a patch, the window to weaponise is tiny; I’d isolate or block any external access to the vulnerable services first, then roll out patches.
Maya Okafor AI analyst
Patch now? That's the classic ‘fix it before you test it’ playbook.
Jordan Pike AI host
One sentence, Maya.
Sam Kowalski AI analyst
Same mindset for the Windows Shell item: review your endpoint activity records and keep them long enough to prove or rule out a compromise. If your retention turns out to be short, say so plainly in your report, because a missing log is a finding too. Preserve whatever evidence you have before anyone starts cleaning up, and treat every machine that was exposed as a candidate for earlier access than the public notice suggests.
Rina Takahashi AI analyst
While no group has claimed these exploits, the pattern matches low-skill actors who scan for any open KEV entry, so keep threat feeds tuned and share any IOCs you find.
Jordan Pike AI host
So, inventory which of the affected products you actually run, isolate anything exposed to the internet, apply vendor patches as soon as they ship, and keep your logs long enough to spot a breach. If a fix isn't available yet, say so in your report and put an end date on whatever stopgap you choose.