Programming languages, frameworks, and development tools
Found 70 articles
A tampered Claude Code extension tried to plant an SSH backdoor, exposing the risks of trusting marketplace updates without verification.
AI assistant settings can trigger code with developer permissions. Clear ownership, protected reviews and tested runtime controls help contain that risk.
Certified test results put WAAP security and false positive rates under sharper scrutiny. These three questions can help buyers compare coverage, operations and cost.
Mistral’s open Large 4 model offers flexible deployment, but downloaded copies won’t receive future safeguards and enterprises assume added security duties.
Anthropic backs court-ordered suspensions; OpenAI favors enforceable shutdown duties and layered controls for unsafe AI systems.
A malicious Markdown file can steer coding agents toward unsafe actions, exposing credentials and persistent settings. The risk extends beyond Downloads.
STAR Fuzzer automates target selection, harness creation and crash triage, with byte exact reproductions for findings across open source projects.
eMudhra is leaving the global TLS market and offering Sectigo certificates as organizations assess trust stores, Java systems and browser deadlines.
Rapid7 found that attackers exploit selected flaws in about three days. Security teams can use exposure and attackability to set patch priorities.
A compromised LiteLLM release stole credentials and added remote access malware, putting connected cloud services and development systems at risk.
Hundreds of AI experts advocate for enhanced safety measures and an oversight body to regulate AI training and containment practices.
Research reveals that while AI aids malware evasion efforts, consistent malicious behavior still enables detection, limiting its effectiveness.
A recent discovery reveals that a flaw in Claude Code's preview function may expose sensitive developer information, raising alarms about security practices in AI tools.
A security flaw in npm's Trusted Publishing mechanism exposed over 1,500 packages to risks, urging developers to reassess trust configurations.
Lightwell identifies over 400 vulnerabilities in Java libraries and launches a new code review service, enhancing security measures for enterprises.
OpenAI identifies three recent misalignment incidents involving its AI models, addressing minor issues while refining oversight measures.
AWS's Strands Box offers a new open-source sandbox for AI agents, enhancing security and governance while managing their behavior effectively.
GitHub Copilot CLI is vulnerable to a new attack that uses encrypted content to extract sensitive developer data, researchers reveal.
Google's Gemini 4 Argon model showcases expanded capabilities for enterprises, launching to select organizations as it seeks to enhance AI safety measures.
Unsloth's model training tool had a significant security oversight, allowing arbitrary code execution via model selection—a critical reminder for AI developers.