AI Risk Wire

Cases › security breach › LR-2026-0781

OPPO Find X10 Manus AI agent email access: indirect prompt injection vulnerability disclosed

Preliminary Why: 1 report: Tech Times · subject to change · 1 independent report

Status: no action reported

Severity negligible
Reported 24 Sep 2026
Developer and model Butterfly Effect

No source states: loss or records, agent tool, could do, deployer, industry.

Severity negligible: Test: no effect beyond the system itself. Fact: the vulnerability was patched and the fix confirmed effective before publication; the source reports no exploitation, affected parties or data loss.

Manus, an AI agent platform built by Butterfly Effect, can act on a user's behalf with access to resources such as email. On September 24, 2026, Salt Labs published a disclosure of an indirect prompt injection vulnerability. An attacker could send an email containing instructions that the agent processes. The embedded instructions, encoded with JSFuck, were reported to direct the agent to execute attacker-controlled code within its virtual environment. The vulnerability was disclosed through Meta's bug bounty program. Salt Labs confirmed the fix was effective before publication. The sources do not report exploitation in the wild, affected users, or data exposed.

What happened

1 · How it was caughtSalt Labs published the vulnerability disclosure on September 24, 2026.

What was done

No source states an action or fix.

Risks

AI used to run a cyber attack
Primary
96 incidents
Rising: 24 vs 4.0 a week

Sorted by rule from the case's fields; not yet read from the sources.

Controls

Not yet read for controls. Controls that mitigate the primary risk:

Model

Safety training and refusal · Cyber and weapons safeguards · Abuse monitoring and account enforcement · Staged or restricted release · Dangerous-capability testing before release · Outside red-teaming · Content safety classifiers · Safeguards kept on in testing · Reasoning monitoring

Infrastructure / target

Known flaws patched · Multi-factor sign-in and password policy · Network segmentation · Intrusion and endpoint detection · Backups and recovery · Phishing and fraud filters

Similar incidents

All cases of this kind ›

Evidence

No source states: what the AI was doing, what went wrong, what it reached or did, when it happened, how many records were exposed, the direct loss, a count of the other parties affected, any court case, the cause, what set it off, what was done.

Show the full account: 1 items from the primary sources

What happened, in order

  1. 2026-09-24 Salt Labs published the Manus prompt injection disclosure.
Show every field, with its source and grade

The AI system

Developer
Butterfly Effect
C one secondary source states it

What happened

Earliest source we hold
2026-09-24
C one secondary source states it
What went wrong
security breach
D inferred by us from the described behaviour, reason shown

Impact

Severity
negligible
D inferred by us from the described behaviour, reason shown
Why this severity
Test: no effect beyond the system itself. Fact: the vulnerability was patched and the fix confirmed effective before publication; the source reports no exploitation, affected parties or data loss.
D inferred by us from the described behaviour, reason shown

Notes on gaps

The article's claims about China's National Intelligence Law, OPPO's ownership, Manus's funding and Meta acquisition are background and are not used as incident facts. Manus's Cue feature is not confirmed as the feature that was exploited.

LR-2026-0781 · version 1 · published 2026-11-01 after automatic checks; not yet checked field by field by a person · JSON