Latest
  1. Johns Hopkins Bayview tells 3,182 patients an employee put their health data into a public AI tool
  2. China’s AI compute access expected to double in 2027, with US-designed chips about 65 per cent
  3. UK commits £100m to national AI Biosecurity Lab
  4. Moonshot AI releases Kimi K3.1 Pro with 1-million-token context window
  5. Residents in Newport and Cardiff oppose hyperscale datacentres near homes
  6. Study: Grok 4.7 picked Republicans in 92% of forced midterm answers, GPT-6.1 Sol picked Democrats in 97%
  7. Attackers exploiting the PHP flaws hand the break-in to AI agents, Unit 42 says
  8. Elvis Presley’s rightsholders sue Anthropic over training on 274 songs

Latest / Frontier models and safety

IncidentFrontierGovernmentInternationalConfirmed

Attackers exploiting the PHP flaws hand the break-in to AI agents, Unit 42 says

See how this story connects

Source: Palo Alto Networks Unit 42Also reported by The Hacker News

Palo Alto Networks’ Unit 42 updated its threat brief on the PHP2Shell flaws (CVE-2026-109618 and CVE-2026-109619) on 23 October to report attackers using agentic AI attack platforms after exploiting them. In one case against a government entity in Europe, the attackers reached password hashes and multi-factor authentication secrets within two hours of the first exploit, set up an NTLM-relay tunnel into Active Directory and tried to gain Domain Admin rights. Unit 42 says the commands came in batches wrapped in delimiters such as ==SHELLUSER== and ==PASSWD== that produce output for an AI model to read, with short serial tasks and quick pivots after failures. It counts more than 22.6 million exposed instances that may be vulnerable.

Why it matters

Flaws found with frontier models are now being exploited by AI agents too, and a government network was compromised within two hours of the first exploit.

When
23 Oct 2026
ID
WTA-N-2026-0423
Who
Palo Alto Networks, Unit 42, PHP Foundation
System
agentic AI attack platforms (unnamed)
Operator
unknown attackers exploiting the PHP2Shell flaws
Type
post-exploitation of a European government entity: password hashes, MFA secrets, NTLM relay into Active Directory
Status
reported 23 Oct 2026; Domain Admin escalation attempted; disclosed by Palo Alto Networks Unit 42 · Incident ledger

Summary by wheresthe.ai, from Palo Alto Networks Unit 42 and the reporting credited above. Checked against the source on 25 Oct 2026.

Read the original at Palo Alto Networks Unit 42

Source copy: download the page as saved on 25 Oct 2026 (PDF). Also held by the Internet Archive.

Next stop on this Storyline: What happened nextRansomware affiliates ran intrusions through AI agents, an exposed server shows

Cite this story

WTA-N-2026-0423: “Attackers exploiting the PHP flaws hand the break-in to AI agents, Unit 42 says.” wheresthe.ai, 23 Oct 2026. https://wheresthe.ai/d/intl-attackers-use-ai-agents-after-php2shell-exploits-2026-10/ (source: Palo Alto Networks Unit 42).

Spotted an error? Email [email protected].