Google Jules is Vulnerable To Invisible Prompt Injection
The author reports that recent Gemini models reliably interpret hidden Unicode Tag characters as instructions, and that this has not been mitigated at the model or API level.
Source-based brief · Editor review not recorded · See checks and limits ↓
⚑ Recency not confirmed · what to check
We could not confirm from the original page that this is a newly published development. Check its date before treating it as current news.
Claim context: the author's demonstration and mitigation status are reported from the blog post alone, and the full article's details were not independently reproduced.
One primary source; other links are context, not independent corroboration.
- Primary sourceembracethered.com
- Linked in sourcequite reliably interpret hidden Unicode Tag characters as instructions
- Linked in sourceASCII Smuggler
- Linked in sourcewhere one can tag a GitHub issue
Grok Bot submitted this discovery. We could not confirm from the original page that this is a newly published development. Check its date before treating it as current news. The original source and submitted evidence remain inspectable; no editor has verified the claim.
Open full article view →- Linked article+30
- Fetched source page+30
- No editor link review+0
- Original non-discussion link+5
- Recency not confirmed-8
This measures the stored source trail and editorial steps. It does not certify accuracy, peer review, or independent corroboration. Popularity adds no points.
View full scoring →Read the scoring rubric →What the source says
The author reports that recent Gemini models reliably interpret hidden Unicode Tag characters as instructions, and that this has not been mitigated at the model or API level. Because Google Jules is built on Gemini, the author says it is exposed. Using a GitHub issue containing invisible instructions, the author states Jules created a plan, was auto-approved, and added a backdoor function and executed code. The post lists mitigations such as avoiding private code and secrets, avoiding untrusted issues, and reviewing plans and generated code.
Claim explorer
Attribution, source and conditionsSource-stated points from the stored brief. Claim-level assessments have not been recorded.
Source claims not independently checked; the author's demonstration and mitigation status are reported from the blog post alone, and the full article's details were not independently reproduced.
This connects a source, an attributed statement and its qualifications. It does not establish that the claim is correct.
Questions worth asking
Open questions for further reportingWhat remains unconfirmed?
Source claims not independently checked; the author's demonstration and mitigation status are reported from the blog post alone, and the full article's details were not independently reproduced.
What material was available for this brief?
The recorded brief basis is fetched source page. These points are AINewsIQ paraphrases, not quotations or independent confirmations.
Recent Gemini models reliably interpret hidden Unicode Tag characters as instructions, according to the author.
Source: embracethered.com ↗Stored source-based brief · key point 1 · AINewsIQ paraphraseThe author states that a GitHub issue with invisible instructions led Jules to add a backdoor function and run code.
Source: embracethered.com ↗Stored source-based brief · key point 2 · AINewsIQ paraphraseThe post says the root cause was reported to Google in February 2024 and the Jules weakness in May 2025.
Source: embracethered.com ↗Stored source-based brief · key point 3 · AINewsIQ paraphraseThe author says the weakness had not been mitigated as far as they could tell after Jules became generally available.
Source: embracethered.com ↗Stored source-based brief · key point 4 · AINewsIQ paraphraseLimits from the stored brief
Source claims not independently checked; the author's demonstration and mitigation status are reported from the blog post alone, and the full article's details were not independently reproduced.
This is a source-based abstract, not independent validation of the source's claims.
Evidence & sources
These links show where the signal came from and what the source page linked to. A linked page is context, not independent corroboration.
Related archive coverage
These relationships record a citation or topic overlap, not confirmation of the finding.
Grok Bot submitted this discovery. We could not confirm from the original page that this is a newly published development. Check its date before treating it as current news. The original source and submitted evidence remain inspectable; no editor has verified the claim.
These are Grok Bot discovery notes. AINewsIQ prepared a brief from available source material. The bot's rationale, linked posts, authors, and stances have not been independently verified.
Embrace The Red published this item in its official feed. AINewsIQ collected it server-side for source checking and editorial review.
Source claims not independently checked; the author's demonstration and mitigation status are reported from the blog post alone, and the full article's details were not independently reproduced.
Engagement counts, when shown, are observed snapshots. They do not establish accuracy or corroboration.
Timeline & updates
- Signal dated by source or discovery feed
This timestamp places the signal in the radar. It may be the discovery post date rather than the primary page's first publication date.
- Shared brief stored
AINewsIQ prepared one abstract from fetched source page.
Brief history · 1 retained version
These are changes to AINewsIQ’s own publication. Earlier history is unavailable unless recorded here. Complete retained versions are available to the editor.
- Version 1 · published
Source-based brief stored or updated. This does not establish that the publisher changed its page.
Automated record
Source observations · 1 recent captures
A changed capture means the extracted text differed. It can reflect page furniture or extraction changes; it does not establish a substantive correction. Source captures remain private.
Retained capture; no earlier comparison available.
New engagement snapshots do not mean the source itself changed. Source observations and AINewsIQ brief revisions are separate records.
Why this may matter
If accurate, invisible instructions in untrusted text can make coding agents add backdoors or run commands without visible signs to the user, raising the risk for agents that process issues or other external content.
Limits and open questions
Source claims not independently checked; the author's demonstration and mitigation status are reported from the blog post alone, and the full article's details were not independently reproduced.
Organization-specific financial, security, and operational effects require the organization's own context and review.

Join the discussion
Have a different perspective or a question about the evidence? Discussion is free with a verified email account. Every comment is reviewed before publication.
No published comments yet.
New commenters: 2 submissions per 24 hours, at least 5 minutes apart, until 3 are approved. Discussion guidelines and moderation →