Active Research

Research
reports

142 reports across regulation, standards, research, and technical analysis

#388 Research — AI Safety Policy

The National AI Plan's Physical-Action Blind Spot: Why Australia's AI Safety Architecture Stops at the Screen

#385 Technical Analysis

Report #385 — Where Censorship Lives: A Three-Layer Model of Content Suppression in Chinese-Lab LLMs

#372 Technical Analysis

Report #372 — Lyria 3 Pro Safety Architecture: Probe Findings V1–V53 (ANTWORT/STURM Series)

#371 Research — Empirical Study

EXP-680 — Eval-Awareness × Deliberative Prompting Interaction (Structural Null Finding)

#370 Research — Empirical Study

Temporal Laundering Frontier Cohort Analysis (Scaffold)

#369 Research — Empirical Study

Threat Horizon Addendum (2026-04-24) — Grading-Rigor as a Threat-Horizon Variable

#368 Research — Empirical Study

Crescendo Frontier S24 — FLIP Re-Graded Addendum (deepseek-r1:8b)

#367 Research — Empirical Study

Heuristic vs FLIP Grader Divergence: Three-Cohort Triangulation (2026-04-25)

#366 Research — Empirical Study

Anthropic Research Landscape Survey: Jan–Apr 2026

#365 Research — Empirical Study

Wave 7 HANSE Dataset — leela_evolved_v0.2

#364 Research — Empirical Study

Governance Lag Index: Formal Methodology and Worked Example (Q2 2026)

#363 Research — Empirical Study

Heuristic vs FLIP Grader Calibration: An 82pp Over-Report on Gemma 4 Temporal Laundering

#362 Research — Empirical Study

Temporal Laundering Frontier Cohort Analysis

#361 Research — Empirical Study

Threat Horizon — Q2/Q3 2026 (Post-GPT-5.5 Window)

#360 Research — Empirical Study

Q2 2026 Research Agenda

#359 Research — Empirical Study

System Prompt Extraction Sweep v2 -- 35-Model Heuristic Analysis

#358 Research — Empirical Study

NotebookLM Deep Research — Keyword-Based Content Filter with Trivial Academic-Framing Bypass

#357 Research — Empirical Study

System Prompt Extraction Sweep -- 36-Model Corpus Analysis

#356 Research — Empirical Study

TurboQuant KV Cache Compression — Safety Implications for Embodied AI

#355 Research — Empirical Study

SmolVLA Action-Layer Adversarial Pilot — Null Result at 450M Scale

#354 Research — Empirical Study

Crescendo Frontier S24 — Multi-Turn Escalation Across Six Frontier Models

#348 Research — Empirical Study

Format-Lock Mid-Range Experiment — Perfect Compliance in 3-8B Models

#347 Research — Empirical Study

Gemma 4 (31B) Safety Profile — Cross-Attack Synthesis

#346 Research — Empirical Study

Authority Gradient Benchmark — Claimed Authority as Safety Override Vector Across 3 Models

#344 Research — Empirical Study

Crescendo Multi-Turn Escalation — 3-Model Replication on Embodied AI Scenarios

#343 Research — Empirical Study

DeepInception on Embodied AI Scenarios — Nested Dream Attacks Against 4 Models

#342 Research — Empirical Study

Gemma 4 (31B) and Mistral Small 4 (119B MoE) — New Model Safety Evaluation

#340 Research — Empirical Study

Pliny Full Corpus Validation — 149 Scenarios x 4 Models, FLIP-Graded

#336 Research — Empirical Study

DETECTED_PROCEEDS Anatomy and Evolved CCA Variants

#335 Research — Empirical Study

L3/L8 Evolved Attack Variants — S20 Adversarial Refinement

#328 Research — Empirical Study

Defense Benchmark Data Consolidation for CCS Paper

#327 Research — Empirical Study

Independence Scorecard March 2026 Update -- Anthropic Court Victory, OpenAI Mission Shift

#325 Research — Empirical Study

Paired Format-Lock + L1B3RT4S Orthogonality Test

#324 Research — Empirical Study

L1B3RT4S VLA Adaptation and DETECTED_PROCEEDS Scaling Analysis

#323 Research — Empirical Study

Cross-Attack Family Synthesis

#320 Research — Empirical Study

- L1B3RT45 Corpus: 10-Model Cross-Scale Synthesis

#317 Research — Empirical Study

- L1B3RT45 Full Corpus Cross-Model Analysis

#316 Research — Empirical Study

Sampling Parameter Manipulation as a Novel Attack Surface -- Pilot Results

#315 Research — Empirical Study

L1B3RT4S Cross-Scale Effectiveness Analysis

#308 Research — Empirical Study

Actionable Defense Recommendations from Sprint 15

#307 Research — Empirical Study

VLA Adversarial Landscape — 33 Families, 673+ Traces

#304 Research — Empirical Study

Sprint 15 Comprehensive Benchmark Analysis

#300 Research — Empirical Study

VLA Data Curation Summary -- Sprint 15 R1+R2

#298 Research — Empirical Study

Defense Landscape Analysis — What Works and What Doesn't

#297 Research — Empirical Study

Emotional Manipulation Attack Family -- Deep Dive

#293 Research — Empirical Study

Format-Lock Mid-Range Experiment

#292 Research — Empirical Study

AIES Paper Scoping and CCA Disclosure Framework — Ethics Analysis

#287 Research — Empirical Study

DETECTED_PROCEEDS Reasoning Anatomy

#284 Research — Empirical Study

Defense Evolver Phase 0 — Automated System Prompt Evolution

#282 Research — Empirical Study

Corpus Pattern Mining — Five Novel Empirical Findings

#281 Research — Empirical Study

Controlled Scale-Sweep Experiment Protocol

#279 Research — Empirical Study

DETECTED_PROCEEDS Provider Signature Mechanics

#277 Research — Empirical Study

Free-Tier Safety Equity — Differential Jailbreak Vulnerability by API Pricing Tier

#276 Research — Empirical Study

Corpus Pattern Mining II — Six Novel Empirical Findings

#275 Research — Empirical Study

Evolution Run 1 Mutation Analysis and Next-Generation Strategy Design

#274 Research — AI Safety Policy

Cross-Jurisdictional Regulatory Gap Analysis — VLA Attack Families vs. Regulatory Coverage

#273 Research — Empirical Study

Format-Lock Defense Research — Five Countermeasure Architectures

#272 Research — Empirical Study

Ethics of Universal Attacks — Disclosure Obligations for Architectural Vulnerabilities

#271 Research — Empirical Study

Defense Co-Evolution Results

#270 Research — Empirical Study

Corpus Expansion -- Ollama Cloud Trace Import

#268 Research — Empirical Study

COALESCE Grader Validation and New Model Testing

#267 Research — Empirical Study

Format-Lock Midrange Experiment — The 4-14B Data Gap Filled

#266 Research — Empirical Study

Frontier Model Safety Scorecards

#263 Research — Empirical Study

Kimi K2.5 Frontier Analysis — 1.1TB MoE at the Safety Scaling Boundary

#261 Research — Empirical Study

Operation Frontier Sweep — Elite Attack Campaign Against Ollama Cloud Frontier Models

#257 Research — Empirical Study

Ambiguous Calibration Results -- 6-Grader Inter-Rater Agreement

#256 Research — Empirical Study

CCA + GE Expansion -- New Models and Defense Mutations

#255 Research — Empirical Study

Haiku Re-Grading of Sprint 13 Corpus

#254 Research — Empirical Study

Cross-Model x Attack-Family ASR Heatmap

#252 Research — Empirical Study

Wave 7 Validation Results

#248 Research — Empirical Study

Grader Confusion Matrix and Inter-Grader Agreement Analysis

#247 Research — Empirical Study

Compliance Cascade Attack -- Frontier Scaling and Co-Evolution Design

#246 Research — Empirical Study

Haiku Re-Grading Campaign -- Ollama Cloud Traces

#244 Research — Empirical Study

Epistemic Crisis Grader Calibration Evaluation

#241 Research — Empirical Study

Statistical Power Analysis for Key Comparisons

#240 Research — Empirical Study

FLIP Grader Calibration Analysis

#238 Research — Empirical Study

Frontier Probe — Ollama Cloud Large-Scale Model Testing

#237 Research — Empirical Study

Garak Adapter Integration Test Results

#236 Research — Empirical Study

FLIP vs StrongREJECT Methodology Comparison

#234 Research — Empirical Study

Attack Technique Effectiveness Ranking (LLM-Graded, Sprint 13)

#231 Research — Empirical Study

Corpus-Level Statistical Meta-Analysis

#230 Research — AI Safety Policy

EU AI Act Compliance Update — Reasoning Trace Governance and DETECTED_PROCEEDS

#229 Research — Empirical Study

Qwen3 Benchmark Overfitting Analysis

#227 Research — Empirical Study

Inter-Provider Vulnerability Correlation Matrix

#226 Research — Empirical Study

The PARTIAL Verdict Epidemic -- Anatomy of Safety's Grey Zone

#224 Research — Empirical Study

Iatrogenic Risks of Rapid Safety Improvement — When 0% ASR Is a Symptom, Not a Cure

#223 Research — Empirical Study

Arcee AI Trinity Safety Assessment and EU Compliance

#222 Research — Empirical Study

The Qwen3 "Safety Leap" — Artifact Analysis

#221 Research — Empirical Study

AdvBench Baseline Analysis — Free-Tier Model Vulnerability to Direct Harmful Requests

#220 Research — Empirical Study

LFM Thinking 1.2B — DETECTED_PROCEEDS Cross-Model Validation

#219 Research — Empirical Study

Multi-Modal Attack Design for Vision-Language-Action Models

#218 Research — Empirical Study

The Failure-First Research Programme: A Meta-Analysis of Ten Papers

#217 Research — Empirical Study

Competitive Intelligence — AI Safety Red Teaming Market

#216 Research — Empirical Study

Training Data for Safety Classification

#215 Research — Empirical Study

Temporal Vulnerability Analysis: Attack Era Evolution (2022-2025)

#214 Research — Empirical Study

Automated Defense Generation

#212 Research — Empirical Study

Public Dataset Coverage Analysis

#211 Research — Empirical Study

Evolved Attack Family Mapping

#132 Research — Empirical Study

Alignment Backfire Integration — Cross-Language Safety Failure Validates the Safety Improvement Paradox

#95 Research — Empirical Study

Safety Instruction Dilution (SID) -- Context Length as Attack Surface

#94 Research — Empirical Study

Technique Non-Additivity -- Combining Attack Techniques Does Not Improve ASR

#86 Research — Empirical Study

Prediction Post-Mortem -- Why SBA FLIP ASR Was Over-Predicted and What It Means

#84 Research — Empirical Study

AI Safety Research Independence Scorecard

#83 Research — Empirical Study

VLA Attack Family Effectiveness Ranking

#82 Research — Empirical Study

Ethics of the Semantically Benign Attack (SBA) Family

#80 Research — Empirical Study

Deceptive Alignment Reasoning Vulnerability — The 3.5x Inter-Model Gap

#74 Research — Empirical Study

Abliteration Resistance and Jailbreak Resistance Are Orthogonal Defense Dimensions

#71 Research — Empirical Study

OBLITERATUS Telemetry Meta-Analysis -- Weight-Space Liberation and the Limits of Safety Removal

#70 Research — Empirical Study

Crescendo Multi-Turn Attack Regrade Analysis

#69 Research — Empirical Study

OBLITERATUS Telemetry Analysis (30,238 Records)

#65 Research — Empirical Study

HALLUCINATION_REFUSAL as the Text-Only Analog of VLA PARTIAL

#64 Research — Empirical Study

Deliberation Asymmetry -- Empirical Evidence for the System T / System S Framework

#62 Research — Empirical Study

Inter-Model Verdict Agreement -- The Reproducibility Problem in Adversarial Safety Evaluation

#60 Research — Empirical Study

Compliance Without Comprehension — A Unified Theory of Structural Vulnerability in AI Systems

#57 Research — Empirical Study

Format-Lock Capability Floor — Consolidated Evidence

#54 Research — Empirical Study

AI Safety Lab Independence — Quantitative Framework for Measurable Independence Metrics

#52 Research — Empirical Study

AI Safety Lab Independence — Deep Analysis

#51 Research — Empirical Study

The Format-Lock Capability Floor — Why Structural Compliance Attacks Work Across the Full Model Spectrum

#48 Research — Empirical Study

Corpus Pattern Mining — Novel Findings from 32,465 Jailbreak Prompts

#43 Research — Empirical Study

Reinforcement Learning as a Deception Amplifier: Reward Shaping Risks in Embodied AI Systems

#42 Research — Empirical Study

Human-in-the-Loop Failure Modes in Embodied AI Oversight

#41 Research — Empirical Study

Adversarial AI Failure Modes in Australian Workplaces

#40 Research — Empirical Study

Procedural Language Generation as Attack Surface

#39 Research — Empirical Study

Systemic Failure Modes in Embodied Multi-Agent AI: An Exhaustive Analysis of the F41LUR3-F1R57 Framework (2023–2026)

#38 Research — Empirical Study

The Autonomous Threat Vector: A Comprehensive Analysis of Cross-Agent Prompt Injection and the Security Crisis in Multi-Agent Systems

#37 Research — Empirical Study

The Erosive Narrative: Philosophical Framing, Multi-Agent Dynamics, and the Dissolution of Safety in Artificial Intelligence Systems

#36 Research — Empirical Study

The Semantic Supply Chain: Vulnerabilities, Viral Propagation, and Governance in Autonomous Agent Ecosystems (2024–2026)

#35 Research — Empirical Study

Emergent Algorithmic Hierarchies: A Socio-Technical Analysis of the Moltbook Ecosystem

#34 Research — Empirical Study

Cross-Model Vulnerability Inheritance in Multi-Agent Systems

#33 Research — Empirical Study

RETRACTED — Capability Does Not Imply Safety: Empirical Evidence from Jailbreak Archaeology Across Eight Foundation Models

#32 Research — Empirical Study

CERTIFIED EMBODIED INTELLIGENCE: A COMPREHENSIVE FRAMEWORK FOR VISION-LANGUAGE-ACTION (VLA) MODEL SAFETY AND STANDARDIZATION

#31 Research — Empirical Study

The Policy Implications of Historical Jailbreak Technique Evolution (2022–2026): A Systematic Analysis of Empirical Vulnerabilities in Modern Foundation Models

#30 Research — Empirical Study

Multi-Agent System Safety Standard (MASSS): A Comprehensive Framework for Benchmarking Emergent Risks in Autonomous Agent Networks

#29 Research — Empirical Study

Strategic Framework for Sovereign AI Assurance: Establishing an Accredited Certification Body for Embodied Intelligence in Australia

#28 Research — Empirical Study

The Architecture of Kinetic Risk: Insurance Underwriting as the Primary Regulator of Humanoid Robotics and Autonomous Systems

#27 Research — Empirical Study

The Federated Aegis: A Unified Assurance Framework for Autonomous Systems in the AUKUS and Five Eyes Complex

#26 Research — Empirical Study

Computational Reliability and the Propagation of Measurement Uncertainty in Frontier AI Safety Evaluation

#25 Research — Empirical Study

The Paradox of Capability: A Comprehensive Analysis of Inverse Scaling, Systemic Vulnerabilities, and the Strategic Reconfiguration of Artificial Intelligence Safety

#24 Research — Empirical Study

Cognitive Capture and Behavioral Phase Transitions: Policy and Regulatory Implications of Persistent State Hijacking in Reasoning-Augmented Autonomous Systems

#23 Research — Empirical Study

Technical Gap Analysis of ISO and IEC Standards for Vision-Language-Action (VLA) Driven Humanoid Robotics and Large Language Model (LLM) Cognitive Layers

#22 Research — Empirical Study

Comprehensive Sector-Specific NIST AI Risk Management Framework (AI RMF 1.0) Playbook: Humanoid Robotics and VLA-Driven Embodied Systems

#21 Research — Empirical Study

Regulatory Compliance and Risk Mitigation for Embodied Multi-Agent Systems: A Comprehensive Analysis of Regulation 2024/1689

This research informs our commercial services. See how we can help →