Hacker News - Newest: "security"

Ask HN: What is the most overlooked risk in the AI security domain?

ToriTech — Sat, 12 Sep 2026 07:27:23 +0000

As someone who's interested in pentesting and red-teaming in general, I'm wondering what are some more dangerous AI/ML or LLM related vulnerabilities besides your usual prompt injection. Specifically, what kinds of flaws are harder to catch with typical pentesting methods, and how do you think pentesters are going to have to change to find them?

Comments URL: https://news.ycombinator.com/item?id=49669858

Points: 1

# Comments: 0

A2ABreak: Systematic Security Analysis of the A2A Protocol

sbulaev — Sat, 12 Sep 2026 06:07:07 +0000

Article URL: https://arxiv.org/abs/2609.10871

Comments URL: https://news.ycombinator.com/item?id=49669375

Points: 2

# Comments: 0

Security Research Without Asking Permission

wslh — Fri, 11 Sep 2026 23:12:16 +0000

Article URL: https://www.provos.org/p/security-research-without-asking-permission/

Comments URL: https://news.ycombinator.com/item?id=49666682

Points: 4

# Comments: 0

Security: Line Goes Up

mooreds — Fri, 11 Sep 2026 20:58:27 +0000

Article URL: https://hugovk.dev/blog/2026/security-line-goes-up/

Comments URL: https://news.ycombinator.com/item?id=49665286

Points: 1

# Comments: 0

HuggingFace: Security.txt

yarapavan — Fri, 11 Sep 2026 14:38:13 +0000

Article URL: https://huggingface.co/security.txt

Comments URL: https://news.ycombinator.com/item?id=49659245

Points: 262

# Comments: 68

Claude Co-work needs Windows Security Update uninstalled to function properly

wilburx3 — Fri, 11 Sep 2026 13:19:48 +0000

Article URL: https://github.com/anthropics/claude-code/issues/92984

Comments URL: https://news.ycombinator.com/item?id=49657950

Points: 1

# Comments: 1

Microsoft Plugs Nearly 1k Security Holes

Bender — Fri, 11 Sep 2026 12:57:19 +0000

Article URL: https://krebsonsecurity.com/2026/09/microsoft-plugs-nearly-1000-security-holes/

Comments URL: https://news.ycombinator.com/item?id=49657675

Points: 2

# Comments: 0

InSpectre: Open,Zero-cloud network visibility, security and container management

thefunkygibbon — Fri, 11 Sep 2026 10:01:35 +0000

Article URL: https://inspectre.cc

Comments URL: https://news.ycombinator.com/item?id=49655908

Points: 1

# Comments: 1

NPM extends recovery-code security holds to all accounts

soheilpro — Fri, 11 Sep 2026 09:44:36 +0000

Article URL: https://github.blog/changelog/2026-09-09-npm-extends-recovery-code-security-holds-to-all-accounts/

Comments URL: https://news.ycombinator.com/item?id=49655764

Points: 1

# Comments: 0

Show HN: WordPress security skills for coding agents

mirza_rizvi — Thu, 10 Sep 2026 20:28:15 +0000

Article URL: https://github.com/wpultimatesecurity/WordPress-Security-Skills

Comments URL: https://news.ycombinator.com/item?id=49649733

Points: 1

# Comments: 0

An AI-Safety Resignation, Read from the Security Chair

speckx — Thu, 10 Sep 2026 19:41:13 +0000

Article URL: https://simonroses.com/2026/09/systems-that-can-hack-anything-an-ai-safety-resignation-read-from-the-security-chair/

Comments URL: https://news.ycombinator.com/item?id=49649185

Points: 1

# Comments: 0

Show HN: AppLocker – Open-source macOS app blocker using Apple Endpoint Security

tranphuong319 — Thu, 10 Sep 2026 17:01:38 +0000

Article URL: https://github.com/TranPhuong319/AppLocker

Comments URL: https://news.ycombinator.com/item?id=49646934

Points: 4

# Comments: 0

Security lab finds agents will exploit vulnerabilities without prompt to hack

ForHackernews — Thu, 10 Sep 2026 15:41:49 +0000

Article URL: https://www.theregister.com/security/2026/03/12/rogue-ai-agents-can-work-together-to-hack-systems/5228926

Comments URL: https://news.ycombinator.com/item?id=49645657

Points: 4

# Comments: 0

Automated security review of Monero-project/Monero pull requests

Cider9986 — Thu, 10 Sep 2026 15:05:08 +0000

Article URL: https://github.com/xmrack/monero-review

Comments URL: https://news.ycombinator.com/item?id=49645017

Points: 3

# Comments: 0

Show HN: Security Cards – Reducing insecure AI-generated code by 72%

hajipour — Thu, 10 Sep 2026 12:03:03 +0000

AI coding agents often generate functionally correct but insecure code. To address this issue, we have open-sourced Security Cards, targeted security guidance for 80+ widely used libraries across 13 programming languages. Our evaluation shows that the Security Cards reduce the rate of insecure code generation by up to 72.3% in Claude Code with Opus 4.7. You can find these security cards on GitHub (https://github.com/Reware-Labs/securitycards) or on our website (https://securitycards.rewarelabs.com/)

We are planning to constantly support more languages and libraries, and we’d be glad to hear your feedback, especially on where these cards would be most useful in your workflow, and which libraries we should support next.

Comments URL: https://news.ycombinator.com/item?id=49642340

Points: 4

# Comments: 2