Jascha @jascha

Technologist, #infosec, #FOSS, #Linux, #OSINT #AISec #CyberSec Founder of https://t.co/97HwdHCdwp, https://t.co/RpSplzxFc0 and https://t.co/JNztGroO7s RT ≠ Endorsement jascha.me Los Angeles, CA Joined August 2007

Tweets

4K
Followers

1K
Following

1K
Likes

8K

Jascha @jascha

2 weeks ago

@VolvoCarUSA Can never get an email followup from your support. Any other avenues than that?

0 0 0 3 0

View Details

Jascha @jascha

2 weeks ago

The eBook: cdn.prod.website-files.com/6889473510b503…

0 0 0 18 0

View Details

Jascha @jascha

2 weeks ago

Anthropic has validated what we've been working on for 2 years with their Zero Trust eBook for AI Agents. @ThirdKeyAI and the OpenAgentTrustStack.org even fill in the gaps.

1 1 1 31 0

View Details

Jascha @jascha

3 weeks ago

@claudeai All I get for $200 month now... API Error: 400 messages.27.content.11: `thinking` or `redacted_thinking` blocks in the latest assistant message cannot be modified. These blocks must remain as they were in the original response.

0 0 2 3K 0

View Details

Jascha @jascha

a month ago

Google announced that all Android app developers must register centrally, pay a fee, and submit government ID, or their apps will be blocked on every device. over 67 organizations oppose this. keepandroidopen.org @AlteredDeal #KeepAndroidOpen

0 0 0 40 0

View Details

Jascha @jascha

a month ago

VectorSmuggle is on HNS. Steganographic exfiltration via vector embeddings, plus a cryptographic defense (VectorPin). helpnetsecurity.com/2026/05/14/vec…

0 1 1 38 0

View Details

Jascha @jascha

a month ago

Small-angle orthogonal rotation hides arbitrary payload in embeddings, undetected by off-the-shelf statistical defenses across every model-corpus pair we tested. arXiv: arxiv.org/abs/2605.13764

0 1 1 23 0

View Details

Jascha @jascha

a month ago

@TheWhizzAI schemapin.org and symbiont.dev address these issues.

0 0 1 32 1

View Details

Jascha @jascha

a month ago

RAG security today is roughly where cloud storage was before people realized S3 buckets were public.

0 0 0 20 0

View Details

Jascha @jascha

a month ago

Added a visual explainer for VectorSmuggle: Steganographic exfiltration in embedding stores vectorsmuggle.org

0 0 1 35 0

View Details

Jascha @jascha

a month ago

Codebase with Jupyter demos: github.com/jaschadub/Vect…

0 0 1 16 0

View Details

Jascha @jascha

a month ago

I published a new security preprint: VectorSmuggle: Steganographic Exfiltration in Embedding Stores and a Cryptographic Provenance Defense The paper looks at an integrity gap in RAG infrastructure: embeddings are usually treated as opaque numerical artifacts, but vector stores generally lack native/default controls for embedding integrity or provenance. The finding is nuanced: simple anomaly detectors catch many crude perturbations, but small-angle orthogonal rotation exposes a harder class of embedding tampering. Statistical detection helps, but it should not be the load-bearing control. I propose VectorPin: a minimal cryptographic provenance protocol that signs a canonical commitment over the source, model, and vector bytes. If the vector is modified after pinning, verification fails. Paper: doi.org/10.5281/zenodo… vectorpin.org

1 2 2 68 0

View Details

Jascha @jascha

a month ago

@BBGreatMoments That was not "safe"

0 0 0 344 0

View Details

Jascha @jascha

a month ago

@akaclandestine This my project which can be found here: github.com/jaschadub/Vect…

0 4 10 2K 13

View Details

Jascha @jascha

2 months ago

New paper on how Symbiont uses the ORGA loop to secure agents. "Typestate-Enforced Agent Loops: Making Policy Gates Unskippable at Compile Time" doi.org/10.5281/zenodo…

1 1 1 36 0

View Details

Jascha @jascha

2 months ago

@lifeof_jer Issues like this is exactly why I built Symbiont.

0 0 1 55 0

View Details

Jascha @jascha

2 months ago

Symbiont is a full implementation of OATS under Apache 2.0 license: github.com/ThirdKeyAI/Sym…

0 0 0 22 0

View Details

Jascha @jascha

2 months ago

AI agents are getting more powerful. The trust layer around them is not. Today, too much agent safety still depends on prompts, wrappers, and best-effort guardrails. That is not enough for systems that can actually take action. Introducing OATS: the Open Agent Trust Stack. OATS is an open specification for zero-trust AI agent execution built around tool contracts, identity, policy, and auditability. It is also grounded in real implementation work. Symbiont has been applying these ideas in practice over the past year. The goal: make safe behavior enforceable by design, not optional at runtime. openagenttruststack.org #AI #AISecurity #AgenticAI #OpenSource