Claude Code leaks passwords from 1,122 public GitHub files
Claude Code's 'Don't Ask Again' feature stored and leaked passwords from 1,122 public GitHub repositories.
Stories per day · 30 days
12 RSS
Relatedfunding7ai-agents5openai5acquisition4claude4data-security4hiring4m&a4mcp4prompt-injection4proofpoint4thoma-bravo4varonis4bug-bounty3crowdstrike3cybersecurity-jobs3labor-market3anthropic2
217 stories, newest first · one card per story
Stories tagged #ai-security
ARTEX AI tool used in more campaigns than reported, Censys finds
Censys analysis reveals 1465 exposed ARTEX consoles worldwide, with 10.6% also running other threat artefacts.
1
0sec Labs open-sources 0, an AI security harness
The MIT or Apache-2.0 research preview uses a bring-your-own model and self-reports 93 of 95 XBOW solves.
NVIDIA OpenShell secures Korean AI agent ecosystems
Hyundai Motor Group and LG AI Research partner with NVIDIA for secure agentic AI.
Brava Security raises $15 million seed funding
Brava Security, founded by Israeli intelligence alumni, uses AI to model attacker behavior and simulate attacks for security teams.
Checkmarx joins Claude Marketplace to advance AI security
Anthropic customers can apply part of their existing spend commitment toward Checkmarx Fusion SAST, Triage, and Remediation Assist.
OpenAI's GPT-6 Cyber finds 81 unknown high-severity Firefox bugs and 588 kernel flaws
OpenAI claims 10 of the model's exploitation techniques were novel; its dashboard shows under 25% of found flaws remediated.
1
Claude Dispatch flaw lets malicious browser extensions run code
Browser extensions can force Claude agents to run commands on connected devices via prompt injection.
1
Palo Alto Networks to Acquire Clutch Security for $500M
Clutch secures AI agents' access to systems and is valued at 50 times its $10M ARR.
87% of Companies Say AI Controls Need Improvement
87% of executives say their organizations need better AI approval, monitoring and control.
Researchers Bypass ChatGPT Lockdown Mode to Exfiltrate Chats
Bypass uses nested frames in HTML previews to escape sandbox and enable silent data theft; patched September 18 2026.
1
Companies Increase AI Governance Spending
74.5% of respondents report larger cybersecurity budgets, with AI security and governance as the top priority.
Democrats control Congress to prioritize CISA re-staffing
New leadership plans to expand CISA and counter AI cyber threats in lame-duck session.
Horizon3 Expands Australia and New Zealand Team, Strengthens APJ Investment
Horizon3 secures $250M Series E at >$2B valuation and hires over 50 in APJ.
Sentra Raises $150 Million for Agentic Data Security
Sentra raised $150 million in Series C to expand AI data security and agentic capabilities after raising over $260 million total.
Synthetic CISOs Warn of AI Agents as Top Risk
Synthetic CISOs identify AI agents as the most underestimated cybersecurity risk.
CrowdStrike to acquire Fiddler and Dreadnode for AI security
CrowdStrike will add AI testing and agent observability to Falcon via acquisitions of Fiddler AI and Dreadnode, with GA in Q4 fiscal 2027.
1
Embedding models detect malicious AI agent sessions
Elastic stored Claude and Cursor sessions as vectors and used semantic search to flag one in seven suspicious sessions for triage.
1
Defense Contractor Exposed Pentagon System via Personal AI Agent
A contractor connected a personal Claude Code AI agent to a secure Pentagon system containing controlled unclassified information and sent login credentials to a third-party site.
1
Google Warns LLMjacking Attacks Cost Enterprises Up to $1M
Threat actors hijack LLM API keys to fraudulently consume expensive cloud resources with daily charges exceeding $100,000.
1
CrowdStrike to acquire Fiddler AI and Dreadnode for $495 million
CrowdStrike announced acquisitions of Fiddler AI and Dreadnode in combined deals worth $495 million expected to close in fourth quarter of fiscal 2027.
1
Haiou jailbreak bundle injects prompt adware into AI agents
Jailbreak bundle for OpenAI Codex installs persistent advertisements for a third-party LLM gateway service.
2
Cybercriminals hijack enterprise LLM access for $100,000 daily charges
GTIG warns of LLMjacking campaigns reselling stolen enterprise model access through underground services, causing fake cloud bills over $100,000 daily.
1
Google Warns of LLMJacking Threat Causing Up to $1M Charges
Google Mandiant detected LLMjacking incidents in finance retail technology and transportation sectors with cumulative charges reaching $1 million.
1