AI Safety & Ethics

The Cyberspace Administration’s annual ‘Qinglang’ campaign arrives in a materially different regulatory environment to last year’s edition, and in the same week

The AGCM accepted binding commitments from all three chatbot providers, establishing a concrete benchmark for what ‘adequate’ hallucination transparency must lo

Most AI risks won’t look like attacks. They’ll look like normal work happening without control. Continue reading on Medium »

Lawmakers and AI makers are battling over the wording of new AI laws. The crux is responsibility and accountability. An AI Insider analysis and scoop.

A practical guide to AI-specific threat modeling, vulnerability assessment, and the...

Let’s cut through the hype. AI today can: Write production-ready code Summarize complex research papers Act like a domain expert in seconds And yet… It can also

arXiv:2604.24966v1 Announce Type: cross Abstract: Frontier AI companies first deploy their most advanced models internally, for weeks or months of safety testin

arXiv:2604.25922v1 Announce Type: cross Abstract: We present DenialBench, a systematic benchmark measuring consciousness denial behaviors across 115 large langu

arXiv:2604.25982v1 Announce Type: cross Abstract: Frontier AI both amplifies existing risks and introduces qualitatively novel challenges. Not only is there a n

Securing the Future in the Age of Intelligent Systems Continue reading on Medium »

The current debate over AI Alignment is stuck in a false binary: either AI is fed a list of “thou shalt nots,” knowing it will eventually… Continue reading on M

How India Is Building a Radar That Sees the Invisible Continue reading on The Gravity »

AI isn’t causing a new cognitive collapse. It’s putting a much older habit on an industrial scale: our willingness to hand our judgment… Continue reading on Lat

The Strait of Hormuz is exposing space warfare's future. Resilience comes from scale-deploy minimum viable capabilities now, leverage markets, prioritize speed

Encryption is often treated as the final line of defense. If data is encrypted properly, the assumption is that it cannot influence… Continue reading on Medium

There is no 6 Nimmt! champion, but a $12 domain registration and one Wikipedia edit convinced several bots there was Unlike search engines that let you judge co

The cybersecurity landscape is undergoing a seismic shift. In 2025, artificial intelligence is no longer just a defensive tool — it has… Continue reading on Med

The cybersecurity landscape is undergoing a seismic shift. In 2025, artificial intelligence is no longer just a defensive tool — it has… Continue reading on Med

The Discord Sleuths: A Cautionary Tale of AI Security Vulnerabilities Imagine walking into a high-security research facility, where the world's most advanced ar

When people talk about Artificial Intelligence, the conversation rarely stays practical for long. Continue reading on Medium »

AI risk isn’t where you think. It’s already inside your organization. Continue reading on Medium »

Before data reaches a vector store, training job, prompt log, or agent workflow, it should be classified, authorized, and bound to policy. Continue reading on M

AI systems are being deployed faster than ever. But there’s a problem most teams aren’t talking...

AI and Text-Based Religion Continue reading on Medium »

A $2.275 million settlement exposed what happens when AI developers treat fairness as someone else's problem. I've spent years building the Continue reading on

Australia still does not have a single economy-wide AI Act. Continue reading on The Beacon & Stone Review »

Most enterprise AI pilots are not failing because the models are bad. They are failing because nobody on the security team will sign off... Continue reading on

Demographic Risk: behind the code lies a culture: young, male, fearless. What does that bring to AI safety? Continue reading on Medium »

The judge also warned Musk and Sam Altman to curb their “propensity to use social media to make things worse outside the courtroom” after both sides traded atta

AI loneliness at work is already here. The real question isn’t whether AI is replacing humans. It’s what you’re handing over without… Continue reading on Medium

Warum KI nicht nur schneller, sondern verständiger eingesetzt werden muss Continue reading on Medium »

Are there things AI should never simulate? Continue reading on Medium »

There’s a conversation happening right now that would have deeply unsettled Marcus Aurelius. Continue reading on Medium »

Investigating the $2.1M synthetic identity breach The recent exposure of Emily Hart—a synthetic persona that managed to raise $2.1M for AI startups while amassi

Blast-radius control tells you how much to break. Intent tells you what breaking it will teach. Only one of these has mature tooling. The post The Next Frontier

Continue reading on ILLUMINATION Local News and Documentary »

The most dangerous failures originate at a single point, propagate until they look systemic, and are missed because monitoring watches at the wrong resolution.

Pourquoi l’attentisme français face à l’intelligence artificielle générative pourrait coûter plus cher que l’erreur inverse Continue reading on Medium »

People are depending too much on AI, and it's changing their cognitive abilities. So, I started...

In any busy warehouse or factory, the most dangerous risks are often the ones you can’t see. Management usually focuses on visible… Continue reading on Medium »

5 wins, 0 losses for an unexpected model. 38% of the catalog returns 404. And the highest-impact change wasn’t a model swap. Continue reading on Medium »

The financial sector has been sprinting toward AI adoption. Now, a major EU regulator has raised the alarm that nobody wanted to hear. Continue reading on AI Ad

Why the AI debate keeps missing what spiritual traditions already knew Continue reading on Medium »

Manchetes como “Fotos e textos ajudam a treinar IAs, mesmo sem permissão de usuários“¹, “Caiu na net: mulher teve foto íntima no banheiro… Continue reading on M

Essay 3 in The Trust & Entrustment in AI and Autonomous Systems Series Continue reading on Medium »

A retired neuroscientist refused cancer treatment for over ten months because a chatbot told him the drugs would do more harm than good. Continue reading on Med

What happens when technology looks intelligent… but isn’t ready to be relied on. Continue reading on Pen With Paper »

90% of defenses fail in minutes. Multi-layer architecture that stopped 45 attacks with zero bypasses across regulated industries. Continue reading on Towards AI