#security News & Analysis

Margin Play: A Multi-Agent System For Public Policy Analysis In The Brazilian Equatorial Margin

Learning Discriminative and Generalizable Anomaly Detector for Dynamic Graph with Limited Supervision

Researchers propose a novel framework for detecting anomalies in dynamic graphs using limited labeled data, combining residual representation encoding with a bi-boundary optimization strategy to balance discrimination and generalization. The model-agnostic approach addresses the gap between unsupervised methods (which produce ambiguous boundaries) and semi-supervised methods (which overfit to limited anomalies).

HongCoin investors recover $2M in locked ETH after nine years

HongCoin investors successfully recovered $2 million in locked Ethereum after nine years, highlighting critical vulnerabilities in legacy smart contracts. The recovery underscores the necessity for developers to update older blockchain infrastructure to prevent fund lockups and protect investor assets from technical obsolescence.

Coldcard MK5 ships with 5 major wallet upgrades

Coinkite released the Coldcard MK5, marking the first significant hardware upgrade to its flagship Bitcoin wallet since 2022, featuring five major wallet improvements. The upgrade reflects ongoing competition in the hardware wallet space and Coinkite's commitment to enhancing security and user experience for self-custodial Bitcoin holders.

Coinkite Launches Coldcard MK5: Major UX Upgrades to Flagship Bitcoin Hardware Wallet

Coinkite has released the Coldcard MK5, the latest iteration of its flagship Bitcoin hardware wallet, featuring a larger Gorilla Glass screen, redesigned buttons, and improved NFC capabilities while maintaining the MK4's dual secure element architecture. The upgrade prioritizes user experience improvements for Bitcoin transaction management and custody.

Vyper developers unveil vyupgrade tool for safe contract rewrites

Vyper developers have released vyupgrade, a tool designed to automate safe smart contract rewrites and enhance security. However, the tool cannot detect original logic flaws, requiring developers to conduct thorough audits independently to ensure contract reliability.

OpenAI’s ChatGPT Now Offers Bank Account Integration — Is It Safe?

OpenAI has introduced bank account integration for ChatGPT Pro users through Plaid, enabling budget tracking and financial advice features. The development raises critical questions about data security and privacy implications when AI systems gain access to sensitive financial information.

OpenAI supplies new model to Japan megabanks for security

OpenAI has deployed a new AI security model to Japan's major banks, marking significant institutional adoption of advanced AI for cybersecurity purposes. While this demonstrates AI's potential to strengthen financial security infrastructure, it simultaneously raises concerns about potential misuse and the concentration of powerful AI tools in critical financial systems.

AgensFlow: A Coordination-Policy Substrate for Multi-Agent Systems

AgensFlow is an open-source framework that treats multi-agent LLM coordination as a learnable policy problem rather than a fixed pipeline, enabling dynamic routing decisions across skill protocols, agent roles, and model bindings. Evaluated on distributed systems and security tasks, the framework demonstrates that learned coordination outperforms static designs while reducing exploration costs through warm-started policy graphs.

'No Airdrop': XRPL Developer Repeats Crucial Warning to XRP Community

An XRPL developer has reiterated warnings to the XRP community about the absence of an airdrop, echoing similar cautions from Ripple leadership earlier this month. The repeated warning suggests ongoing confusion or misinformation within the community that requires clarification from official sources.

Polymarket Vice President Denies Contract Hack, Confirms User Funds Are Safe

Polymarket VP Josh Stevens has denied that a contract hack occurred on the platform, asserting that all user funds remain secure despite a $520,000 legacy key compromise. The statement aims to quell speculation about the platform's security and reassure users of the integrity of their holdings.

A shallow dive into formal verification

This article provides an introductory overview of formal verification, a mathematical approach to proving software correctness that has become increasingly important in cryptocurrency and blockchain development. The piece examines how formal verification methods can enhance security and reliability in smart contracts and critical systems.

Amplification to Synthesis: A Comparative Analysis of Cognitive Operations Before and After Generative AI

AI Safety Landscape for Large Language Models: Taxonomy, State-of-the-art, and Future Directions

DeePen: Penetration Testing for Audio Deepfake Detection

Grid-Orch: An LLM-Powered Orchestrator for Distribution Grid Simulation and Analytics

Large Language Models for Agentic NetOps and AIOps: Architectures, Evaluation, and Safety

ChipMATE: Multi-Agent Training via Reinforcement Learning for Enhanced RTL Generation

Language-Based Agent Control

Do Androids Dream of Breaking the Game? Systematically Auditing AI Agent Benchmarks with BenchJack

Unweighted ranking for value-based decision making with uncertainty

Agentic Interpretation: Lattice-Structured Evidence for LLM-Based Program Analysis

Crypto wallet provider Ledger puts U.S. IPO plans on hold due to market conditions

IPI-proxy: An Intercepting Proxy for Red-Teaming Web-Browsing AI Agents Against Indirect Prompt Injection

FLARE: Adaptive Multi-Dimensional Reputation for Robust Client Reliability in Federated Learning