Helsing unveils Clarity, AI software that prepares military target recommendations
Clarity fuses sensor and open-source data into high-confidence recommendations, with a named human required to authorize any weapons handoff.
1
Stories per day · 30 days
2 RSS
Relatedapt283hookedge3nato3espionage2agent1ai-funding1attention-heads1backdoor1clustering1credential-theft1cryptanalysis1disruptive1extraction1face-recognition1fancy-bear1gru1helsing1human-oversight1
33 stories, newest first · one card per story
Stories tagged #defense
AttentionSwitch: Joint Learning for LLM Jailbreak Defense
AttentionSwitch jointly optimizes sparse attention head selection and query-conditioned scaling to defend LLMs against jailbreaks without catastrophic forgetting.
Adaptive clustering breaks TraintoDefend defense in model extraction
SimBreak raises signature recovery from 0% to 100% in protected networks.
Three defenses harden face recognition
Three tricks detect poisoned identities at training and filter inputs at inference.
StateFence Defends Blue-Team LLM Agents
StateFence preserves task state in SOC LLM agents against evidence-chain contamination from indirect prompt injections.
Disruptive raises $10B fund for late-stage AI and defense bets
The Dallas firm’s first blind-pool vehicle will write checks of up to $1 billion after at least $7.5 billion in commitments.
APT28 uses Hookedge and an SSH tunnel against NATO defense firms
A naval shipbuilder intrusion starting July 24, 2026 lasted at least two months and included stolen credentials.
APT28 targeted NATO shipbuilder, engine maker and police agency
Symantec says the operators used Hookedge and an undocumented loader, ld.exe, whose decrypted payload was not recovered.
APT28 targeted European defense firms in two NATO countries, Symantec says
A July Word attachment led to Hookedge and an in-memory tool; activity ran from July 24 to September 24.
1
LTBD: Learnable Trust-Boundary Delimiters for Prompt Injection Defense
LTBD uses learnable trust delimiters to sharply reduce prompt-injection success without changing LLM weights.
AgentTrap: Stateful Feedback Deception against Autonomous Penetration Testing Agents
AgentTrap closed-loop honeypot cut autonomous pentest agents' real-target success from 95.8% to 79.2% and extracted attacker API keys in 18.8% of runs.
SAGE: Similarity-Based Cleaning of Poisoned Training Data from Verified Examples
SAGE flags clean-label training poisons from a handful of expert-verified examples using similarity-weighted prediction.
Selective Channel Restoration for Backdoored Vision-Language Models
Post-training defense restores fragile VLM projector channels, driving backdoor success near zero without inference overhead.
Removing the NEEDLE in the Haystack: Backdoor Removal in LLMs via Weight Orthogonalisation
NEEDLE removes LLM backdoors by weight orthogonalisation with little capability or safety loss.
Region-Level Black-Box Defense Against Stealthy Embedding-Space Backdoors in CLIP
CLIPGuard localizes and inpaints CLIP embedding backdoors, cutting attack success to about 1% in black-box tests.
AEGIS: Audio Endogenous Guarding via Internal Signals Against Large Audio-Language Model Jailbreaks
AEGIS uses mid-layer risk signals to cut audio-model jailbreak unsafe rates from 17.9% to 0.4%.
TraceGuard: Adaptive Multimodal Poison Filtering through Cross-Feature Rank Agreement
TraceGuard filters poisoned image-text training data via rank agreement among six corpus-level features, removing 98.4% of poison and 5.4% of clean examples.
Deception by Design: CISA's Guide to Tricking Cybercriminals
CISA publishes guidance helping resource-limited organizations deploy deception techniques to trap intruders.
Adaptive Traffic Camouflage: Causal and Resource-Aware Defense Against IoT Fingerprinting
A causal IoT traffic-camouflage defense cuts fingerprinting Macro-F1 by up to 43.5% with limited bandwidth cost.
TigerByte Cyber Emerges From Stealth With $3 Million in Funding
TigerByte Cyber emerged from stealth with $3M in seed funding to deliver hardware-enforced cybersecurity for AI and edge devices.
2
Agentic Ransomware: From Human-Operated to AI-Operated Attacks
SOCRadar analyzes the shift from human-operated ransomware to agentic AI-driven attacks and what this transition means for defenders.
Defense Cyber Spending Set to Surge Amid Rising Attacks on Military Systems
MarketsandMarkets projects the cyber warfare market to nearly double from $14.99bn in 2026 to $28.75bn by 2031, driven by attacks on military systems.
1
CounterPersona: Append-Only Defense Against Unauthorized Persona Skill Distillation
CounterPersona appends targeted counter-persona evidence after data collection to block AI systems from distilling an individual's behavioral patterns into reusable skills.
ActGuard: Pre-execution Action Auditing against Indirect Prompt Injection in LLM Agents
ActGuard audits LLM agent actions before execution against predicted tool priors, masking only malicious spans from indirect prompt injections while preserving utility.