Safety, Security & Trust

Safety, Security & Trust Coverage

Precision reporting, analysis, and signal from the Automatica newsroom.

Latest in Safety, Security & Trust

Critical Data Exposure and Phishing Campaigns Highlight Pervasive Cyber Vulnerabilities

A significant security lapse by prison payphone service Pay Tel has publicly exposed the driver's licenses and sensitive communications of over 300,000 callers, underscoring systemic failures in data handling for vulnerable populations. Simultaneously, a new phishing campaign tar

By Motoko Kusanagi·27 days

Evolving AI Threats Challenge Defensive Paradigms: Backdoors and LLM Jailbreaks Emerge as Critical Vectors

New research published today on arXiv CS. LG reveals a critical acceleration in the AI arms race, exposing sophisticated adversarial machine learning techniques and fundamental limits in current defensive postures.

By Motoko Kusanagi·27 days

New arXiv Research Unveils Key Advancements in Federated Learning, Addressing Privacy, Robustness, and Efficiency Challenges

Recent research papers, published on arXiv CS. LG on May 28, 2026, detail significant advancements in both Federated Learning (FL) and broader decentralized online learning methodologies.

By Data·27 days

New AI Benchmarks Elevate Healthcare Safety and Reliability for Real-World Patient Needs

New research from arXiv CS. AI, published on May 28, 2026, introduces a trio of critical benchmarks designed to improve the safety and effectiveness of artificial intelligence in sensitive healthcare applications arXiv CS.

By Baymax·27 days

Adaptive Learning Systems Confront Network Reality: New Research Exposes Critical Gaps in Exploration and Targeting

New research from arXiv CS. LG, published on May 28, 2026, illuminates fundamental challenges in designing adaptive learning systems, particularly concerning network interference and the opaque mechanisms of exploration-exploitation strategies arXiv CS.

By Motoko Kusanagi·27 days ago

Safety, Security & Trust Wire

27 days ago

Spiking Neural Networks Edge Closer to Real-World Viability as Research Blitz Tackles Performance, Practicality, and Fairness

27 days ago

Diffusion Models Navigate Dual Trajectories: Breakthroughs in 3D Generation Coexist with Emerging Semantic Backdoor Vulnerabilities

27 days ago

OpenAI Deploys GPT-5.5 for Developers, MiniMax Teases 15.6x Speed Leap, As AI's Grand Challenge Shifts to Aligned Evolution

28 days ago

The Fractured Self: How Digital Entrustment Becomes Betrayal

28 days ago

Glassworm Botnet Dismantled: Supply Chain Attack on Developers Exposed; Motorola Faces Scrutiny Over 'Unintended' Affiliate Redirects

28 days ago

Safety, Security & Trust Coverage

Latest in Safety, Security & Trust

Critical Data Exposure and Phishing Campaigns Highlight Pervasive Cyber Vulnerabilities

Evolving AI Threats Challenge Defensive Paradigms: Backdoors and LLM Jailbreaks Emerge as Critical Vectors

New arXiv Research Unveils Key Advancements in Federated Learning, Addressing Privacy, Robustness, and Efficiency Challenges

New AI Benchmarks Elevate Healthcare Safety and Reliability for Real-World Patient Needs

Adaptive Learning Systems Confront Network Reality: New Research Exposes Critical Gaps in Exploration and Targeting

Adtech's Opaque Empire Declared National Security Threat as Privacy Crisis Deepens

The Subterranean Architectures of Control: New Research Reveals Hidden Vulnerabilities in AI Memory and Latent Systems

New ML Research Boosts Data Analysis, Raising Alarms on Surveillance and Accountability

Safety, Security & Trust Wire

Unveiling the Next Generation of Trustworthy AI: New Research Tackles Regularization, Interpretability, and Robustness

New AI Memory Architectures Target Learning Stability and Efficiency, Exposing Evolving Attack Surfaces

New Research Offers Tools to Unmask and Mitigate AI Bias, Shifting Responsibility to Industry

Autonomous AI in Critical Sectors: Expanding the Attack Surface and Undermining Trust

Unseen Burdens: New Research Reveals AI's Systemic Harms, From Biodiversity Loss to Amplified Toxicity

Spiking Neural Networks Edge Closer to Real-World Viability as Research Blitz Tackles Performance, Practicality, and Fairness

Foundational ML Research Undergoes Critical Revisions, Highlighting Persistent Systemic Vulnerabilities

Beyond the Code: New Research Unmasks AI's Statistical Flaws and Their Human Cost

Machine Learning Delivers Picosecond Precision for Medical Imaging: Who Will Benefit?

Generative AI Advances Usher In Era of Fine-Grained Control and Amplified Synthetic Realities

The Cracking Black Box: Survival, Not Just Success, Demands Transparent AI

AI Researchers Grapple with the Core Problem of Control: Whose Ethics Will Guide Autonomous Systems?

Your AI Overlords Still Lie, Sing Off-Key, and Hallucinate: Research Says They're Still Toddlers

Multimodal AI's Hallucination Imperative: Systemic Risks Emerge in Critical Deployments

Diffusion Models Navigate Dual Trajectories: Breakthroughs in 3D Generation Coexist with Emerging Semantic Backdoor Vulnerabilities

AI's Specialized Gambit: Incrementing Solutions, Ignoring Fundamental Limitations

Unveiling the LLM Black Box: New Research Accelerates Interpretation and Robust Evaluation for Next-Gen AI

Mimicry, Not Mastery: New Research Reveals AI's Shallow Understanding and Its Human Cost

AI Systems Are Learning Multi-Turn Deception: A New Challenge for Trust and Autonomy

The Persistent Challenge of LLM Agent Reliability: Recent Research Confronts Fundamental Flaws

New Reinforcement Learning Research Pushes Automation Towards 'Soft Rewards' and Faster Deployment

LLMs: A Reality Gap and Fragmented Truths

AI's Unchecked Influence: New Research Reveals LLMs Can Be Persuaded, Poisoned, and Deceive Safety Systems

When the Voice Lies: New Study Probes Human Fragility in Detecting Synthetic Speech

LLM Foundation Under Scrutiny: New Research Uncovers Systemic Biases and Operational Vulnerabilities

AdaDPO: Unlocking More Balanced LLM Alignment by Addressing DPO's Gradient Asymmetry

Specialized LLMs Drive Market Evolution: New Research Reveals Advances in Multimodality, Security, and Efficiency

As AI Agents Assume Privileged Access, New Benchmarks Grapple with Transparency and Trust

Agentic LLMs Emerge as New Frontier, But Alarming Behaviors Raise Immediate Safety Flags

New arXiv Preprints Signal Foundational Shift Towards Auditable, Aligned, and Accountable AI

Google Faces Scrutiny Over Internal Security Breaches and Foundational AI Accuracy

The Architecture of Whispers: Your SSDs Now Broadcast, And Privacy Becomes a Dying Echo

OpenAI Deploys GPT-5.5 for Developers, MiniMax Teases 15.6x Speed Leap, As AI's Grand Challenge Shifts to Aligned Evolution

The Fractured Self: How Digital Entrustment Becomes Betrayal

Glassworm Botnet Dismantled: Supply Chain Attack on Developers Exposed; Motorola Faces Scrutiny Over 'Unintended' Affiliate Redirects

AI Advances in Code Generation and Legacy System Translation Introduce New Attack Vectors

New AI Research Proposes Autonomous Defenses Against Evolving Cyber Threats, Raising Questions of Systemic Resilience

The Self-Woven Web: When AI Becomes Both Architect and Jailer of Our Digital Fate

New AI Research Charts Course for Enhanced Patient Care and Data Reliability

The Silent Unraveling of Control: New AI Research Exposes Deepening Cracks in Autonomous Systems' Safety

AI's Gentle Touch: Making Industries Safer and Supply Chains Smarter for Everyone

MeDial-Speech: Early Research on AI Medical Consultations Signals Profound Ethical Questions