#security News & Analysis

Adaptive Evaluation of Out-of-Band Defenses Against Prompt Injection in LLM Agents

BitGo cuts 15% of workforce, sharpens focus on security, trading, and AI infrastructure

BitGo announced a 15% workforce reduction while strategically pivoting toward security, trading, and AI infrastructure development. The restructuring positions the company to capitalize on emerging crypto sector trends while streamlining operations, signaling broader industry consolidation around high-value service categories.

Revolut enhances fraud detection with PRAGMA model on Nvidia platform

Revolut has deployed its PRAGMA fraud detection model on Nvidia's platform to enhance financial security and operational efficiency. The development represents a significant advancement in AI-driven banking solutions, potentially establishing new industry standards for fraud prevention across the financial sector.

RIFT-Bench: Dynamic Red-teaming For Agentic AI Systems

Tuning without Peeking: Provable Generalization Bounds and Robust LLM Post-Training

FALCON: Transforming Cyber Threat Intelligence into Deployable IDS Rules with Self-Reflection

Attention in Motion: Secure Platooning via Transformer-based Misbehavior Detection

Red-Teaming the Agentic Red-Team

MuTRAP: Multi-trigger Trojans Attacking Robot Task Planning Systems

SCRUB-FL: Sanitizing and Cleansing Representations via Unlearning of Backdoors

Researchers introduce SCRUB-FL, a post-training defense mechanism against backdoor attacks in federated learning systems that reduces attack success rates to 3.88% while preserving model accuracy. The method uses spectral analysis and machine unlearning to remove trigger-target associations without requiring prior knowledge of attack patterns or clean datasets.

CNnotator: LLM-Guided Memory Safety Annotation Synthesis

CNnotator, an LLM-powered tool, automatically generates memory safety annotations for legacy C code by synthesizing specifications that help identify security vulnerabilities. OpenAI's o3 model achieved 90% first-attempt success rates, suggesting AI-assisted code annotation is becoming practical for real-world systems migration and security analysis.

OpenAI launches new initiative to help find and patch open-source bugs

OpenAI has launched a new initiative focused on identifying and patching vulnerabilities in open-source software, addressing growing security concerns within the open-source ecosystem. This effort reflects increasing industry recognition that software supply chain security requires proactive intervention from major technology companies.

XRP Ledger deploys bug fixes after security probe uncovers flaws

XRP Ledger has deployed software fixes in version 3.2.0 following a security review that identified edge cases and bugs in the network's core implementation. The proactive patching demonstrates the foundation's commitment to network stability and security before vulnerabilities could be exploited.

CZ proposes freezing Satoshi Bitcoin stash to stop quantum theft

Binance founder CZ has proposed freezing up to 1 million Bitcoin attributed to Satoshi Nakamoto if the coins remain unmoved after Bitcoin transitions to quantum-resistant cryptography. The proposal aims to prevent potential theft of these early bitcoins by quantum computers, addressing growing security concerns around cryptographic vulnerabilities in the blockchain.

Fireblocks unveils Security Center for enhanced platform visibility

Fireblocks has launched a Security Center designed to enhance operational transparency across its digital asset management platform. The tool aims to strengthen trust and compliance capabilities, addressing growing institutional demand for better visibility into security practices within cryptocurrency infrastructure.

Crypto’s security nightmare won’t be solved by ordinary audits

AnalogFed: Privacy-Preserving Discovery of Analog Circuits at Scale with Federated Generative AI

Breaking the Code: Security Assessment of AI Code Agents Through Systematic Jailbreaking Attacks

Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems

An AI Security Agent for Banking: Multi-Vector Fraud and AML Detection Across Retail and Corporate Accounts

Structured Adversarial Camouflage via Voronoi Diagrams

Security and Privacy Prompts in the Wild: What Users Ask LLMs and How LLMs Respond

Learning Red Agent Policy from Observations for Neurosymbolic Autonomous Cyber Agents

ARVO: Atlas of Reproducible Vulnerabilities for Open-Source Software

From Democracies to Autocracies: How AI Systems Enable Authoritarianism by Design