#security News & Analysis

Security in LLM-as-a-Judge: A Comprehensive SoK

Architecting Secure AI Agents: Perspectives on System-Level Defenses Against Indirect Prompt Injection Attacks

Worldpay and Global Payments Launch Payments DVN on LayerZero for Cross-Chain Security

Bitcoin, stocks rise, oil slides, after report of Iran's willingness to end conflict

Anthropic’s Claude Code leak reveals autonomous agent tools and unreleased models

AI agents like OpenClaw could drain crypto wallets via ‘malicious skills’: CertiK

A Systematic Taxonomy of Security Vulnerabilities in the OpenClaw AI Agent Framework

Senator Lummis, Cassidy introduce Mined in America Act to support the Strategic Bitcoin Reserve

Ripple CEO Talked About A $13 Trillion Opportunity, But Will XRP Investors Benefit From It?

StarkWare Co-Founder Defends ZK Technology Amid Canton, Ethereum, and Solana Rivalry

Sergey Nazarov Details How Chainlink Economics 2.0 Builds a Virtuous Cycle of Security and Fees

Bitcoin’s Three Unsolved Problems Could Hand Ethereum a Long-Term Structural Advantage

Crypto Giant Bitmain Faces Scrutiny As US Senator Flags Trump Family Ties

Sen. Warren targets Bitmain-Trump family ties in letter to Commerce Secretary Lutnick: Bloomberg

Elizabeth Warren presses Commerce over Bitmain security review

XRP Ledger Gets AI Security Upgrade As Ripple Prepares For Bigger Growth

Crypto Grey Zone Explodes: Why Vietnam’s ONUS Bust Is A Warning To Retail Traders

AIP: Agent Identity Protocol for Verifiable Delegation Across MCP and A2A

Researchers introduce Agent Identity Protocol (AIP) with Invocation-Bound Capability Tokens (IBCTs) to address the lack of authentication in AI agent communications via Model Context Protocol and Agent-to-Agent protocols. The protocol achieved 100% attack rejection rate in testing with minimal performance overhead of 0.086% in real deployments.

Fenbushi co-founder offers bounty to recover $42M lost in 2022 hack

Fenbushi co-founder is offering a bounty to recover $42 million lost in a 2022 hack caused by a compromised seed phrase. Investigators have successfully frozen approximately $1.2 million of the stolen funds as part of ongoing recovery efforts.

Graham Allison: The Iran conflict is marked by uncertainty, military interventions struggle with nation-building, and misconceptions about Iran’s nuclear ambitions persist | All-In Podcast

Graham Allison discusses the uncertainty surrounding the Iran conflict and challenges with military interventions in nation-building on the All-In Podcast. Netanyahu's strategic vision is highlighted as potentially reshaping Middle Eastern security dynamics amid unpredictable geopolitical tensions.

Agent Audit: A Security Analysis System for LLM Agent Applications

Agent-Sentry: Bounding LLM Agents via Execution Provenance

Robust Safety Monitoring of Language Models via Activation Watermarking

CyberGym: Evaluating AI Agents' Real-World Cybersecurity Capabilities at Scale

MCP Security Bench (MSB): Benchmarking Attacks Against Model Context Protocol in LLM Agents