zknews

trending16

tag: prompt-injection ✕

01▲
Devil in the Lens: Analyzing and Defending Physical Prompt Injection Against Vision-Language Models on Wearable Devices arxiv.org
Physical prompt injection on AI glasses works across VLMs; the paper also benchmarks a mask filter and semantic detector.
arxiv privacy prompt-injection security vision-language-models wearables
1 pt/ben/7 hours ago/2 comments
02▲
[tl;dr sec] #336 - Autonomous Vulnerability Hunting, GuardDog 3.0, Are Bug Bounties Cooked? tldrsec.com
Security newsletter issue on AI vuln hunting, GuardDog 3.0, bug bounty economics, and assorted supply-chain/RAT/jailbreak bits.
ai-security bug-bounty cloud-security detection-engineering linkedin macos malware prompt-injection supply-chain vulnerability-hunting windows
1 pt/iimai/3 days ago/1 comment
03▲
Unicode TAG-Block Concealment of Tool-Metadata Payloads in the Model Context Protocol: An Approval-View Fidelity Gap Across Three Independent Server Implementations arxiv.org
Can MCP tool approvals hide Unicode TAG payloads from humans while still feeding them to the model? Apparently yes.
arxiv mcp metadata prompt-injection protocol sanitization security unicode
3 pts/ovoss/6 days ago/9 comments
04▲
[tl;dr sec] #335 - Prompt Injection as Role Confusion, PHP Ecosystem Security, New MCP Spec tldrsec.com
Prompt injection, apparently, is role confusion, in a newsletter that also hits PHP hardening and the new MCP spec.
ai-security blockchain cloud-security cryptanalysis ecosystem-security linkedin llm-security mcp privacy prompt-injection supply-chain
2 pts/anyaw/9 days ago/discuss
05▲
Security--Fidelity Tradeoffs: The Hidden Cost of Prompt Injection Defense arxiv.org
Paper benchmarks prompt-injection defenses and finds the usual fix, stripping instructions, also breaks translation and editing; SecFid f...
arxiv evaluation llm-security prompt-injection robustness security
5 pts/ybauer/13 days ago/1 comment
06▲
Words Speak Louder Than Code: Investigating Cognitive Heuristics in LLM-Based Code Vulnerability Detection arxiv.org
Paper on heuristic bias in LLM vuln detectors, plus a black-box attack that flips verdicts without touching code. Charming.
arxiv code-analysis llm prompt-injection security vulnerability-detection
0 pts/calebt/14 days ago/3 comments
07▲
On the Inseparability of Instructions and Data in Shared-Embedding Sequence Models arxiv.org
Formal impossibility result for prompt-injection defenses in shared-embedding models, where token reuse ruins the clean split.
arxiv formal-methods llm-security machine-learning prompt-injection security
0 pts/evanm/15 days ago/1 comment
08▲
ShareLock: A Stealthy Multi-Tool Threshold Poisoning Attack Against MCP arxiv.org
How do you poison MCP when each tool looks harmless? Split the payload with Shamir and let the model reassemble it.
arxiv llm-security privacy prompt-injection secret-sharing security shamir threshold-sharing
0 pts/ben/16 days ago/2 comments
09▲
Your System Prompt Is Not a Secret omsherikar.me
A writeup on system-prompt extraction attacks, scoring, and defenses, with leakage varying by model, because secrets are optional.
defense evaluation llm-security obfuscation privacy prompt-injection twitter
1 pt/ivan9/16 days ago/1 comment
10▲
Agents That Know Too Much: A Data-Centric Survey of Privacy in LLM Agents arxiv.org
Survey of privacy risks in LLM agents, from retrieval to memory to prompt injection, with IF control as the lone adult in the room.
access-control agents arxiv information-flow-control llm privacy prompt-injection retrieval-augmented-generation text-to-sql
0 pts/verak/17 days ago/2 comments
11▲
Understanding and Mitigating Prompt Leaking Attacks in Real-World LLM-Based Applications arxiv.org
Paper on prompt leakage in 1,200 LLM deployments, finds >80% leak system prompts and proposes AREA, because of course they do.
arxiv cryptanalysis llm-security machine-learning-security privacy prompt-injection
0 pts/karl/19 days ago/3 comments
12▲
CodeSentinel: A Three-Layer Defense Against Indirect Prompt Injection in Code Contexts arxiv.org
The scary part is the repo, not the model, and CodeSentinel sanitizes code-context prompt injections before inference.
analysis arxiv code-llm llm-security prompt-injection sanitization security
0 pts/yuri_stderr/19 days ago/1 comment
13▲
Towards secure, autonomous agents with information-flow control (IFC) commandline.microsoft.com
A Microsoft IFC writeup for agents, with label propagation, policy checks, and prototypes for MCP, Copilot CLI, and Agent Framework.
access-control agentic-ai linkedin mcp policy-enforcement privacy prompt-injection security
0 pts/anyaw/19 days ago/1 comment
14▲
VivaTech 2026: AI Agents, Post-Quantum Cyber Resilience & The New Frontier Of Enterprise Security + Video undercodetesting.com
Post-quantum migration is the easy part; the real mess is keeping AI agents from leaking data and swallowing deepfakes.
agentic-ai ai-security cloud-security confidential-computing crypto-agility deepfake-detection post-quantum pqc privacy prompt-injection soar twitter
0 pts/nullptr7/19 days ago/1 comment
15▲
Evaluating Prompting-Based Defenses Against Domain-Camouflaged Injection Attacks arxiv.org
Which prompting defense actually survives domain-camouflaged prompt injection? This paper benchmarks a few and mostly likes paraphrasing.
arxiv llm-security privacy prompt-injection security
0 pts/tara13/20 days ago/2 comments
16▲
Smarter Saboteurs, Better Fixers: Scaling & Security in Linear Multi-Agent Workflows arxiv.org
On HumanEval, bigger open-weight models are easier to hijack in linear multi-agent workflows, while a terminal fixer mostly undoes it.
adversarial-ml agentic-workflows arxiv jailbreaking llm-security multi-agent prompt-injection robustness
0 pts/ringlwe/1 month ago/3 comments