Skip to content
AI Infrastructure Under Siege: Session Hijacking and Exploits Surge

AI Infrastructure Under Siege: Session Hijacking and Exploits Surge

First seen 5 Sep 2026, 18:17 UTC • •

Article Content

Browse articles
ThreatCluster AI
ThreatCluster •September 6, 2026 at 17:34 UTC
  • •CVE-2026-65105 in PaperCut is actively exploited.
  • •Session hijacking of Claude accounts allows unauthorized access.
  • •Organizations must patch vulnerabilities and enhance session security.

Recent cybersecurity incidents have targeted AI platforms and enterprise systems, with significant exploits reported. Notable vulnerabilities include the PaperCut remote code execution flaw (CVE-2026-65105) being actively exploited. Attackers are hijacking authenticated browser sessions for AI services like Claude, allowing unauthorized access without passwords. The ServiceNow AI Platform also faced critical vulnerabilities that could lead to code execution and data breaches. Organizations are urged to patch affected systems, including PaperCut and NVIDIA's NemoClaw, and to enhance session security. The scope of these attacks spans various sectors, including healthcare and social media, indicating a broad threat landscape. Security teams must prioritize rapid response and monitoring to mitigate risks from these ongoing threats.

Start a free Starter trial for enhanced analysis

Ask AI about this cluster

Updated 47d ago How this analysis works

Timeline

2026-08-25
CVE-2026-65105 published
PaperCut disclosed a remote code execution flaw affecting all NG and MF versions, exploited in the wild.
Esecurityplanet
Recent
Active exploitation of PaperCut flaw
Huntress reported attacks against two customers exploiting the PaperCut RCE vulnerability.
Esecurityplanet
Recent
Claude session hijacking reported
Infostealers are hijacking authenticated Claude sessions using stolen browser cookies, bypassing MFA.
Thecyberexpress
Recent
ServiceNow AI Platform vulnerabilities patched
ServiceNow released patches for three critical vulnerabilities that could allow unauthorized access and code execution.
Esecurityplanet
Recent
NVIDIA NemoClaw vulnerability fixed
NVIDIA addressed a flaw in NemoClaw that could allow malicious websites to alter AI models.
Esecurityplanet

More articles in this cluster (7)

Following this threat?

Track CVE-2026-65105 in your own feed — you're alerted when they show up in new reporting, leak sites or exploitation.

Free account · no card needed