All AI Security & AI Safety Posts
-
AI Security
Semantic Adversarial Attacks: Leaving the Perturbation Budget Behind
A semantic attack changes the lighting, the hair colour or the phrasing. The result is enormous in pixel distance, obviously the same thing to a person, and outside the set of inputs any robustness guarantee was written about.
Read More » -
AI Security
The AI Alignment Problem
The AI alignment problem sits at the core of all future predictions of AI’s safety. It describes the complex challenge of ensuring AI systems act in ways that are beneficial and not harmful to humans, aligning AI goals and decision-making processes with those of humans, no matter how sophisticated or powerful the AI system becomes. Our trust in the future of AI rests on whether…
Read More » -
AI Security
A (Very) Brief History of AI
As early as the mid-19th century, Charles Babbage and Ada Lovelace created the Analytical Engine, a mechanical general-purpose computer. Lovelace is often credited with the idea of a machine that could manipulate symbols in accordance with rules and that it might act upon other than just numbers, touching upon concepts central to AI.
Read More » -
AI Security
When ML Bias Becomes a Security Failure
A demographic differential in false match rate is a demographic differential in exposure to impostors. NIST measured it in 2019 and called it a security concern. The security profession has spent nearly seven years reading the other error rate.
Read More » -
AI Security
Adversarial Attacks on AI: What Has Actually Been Demonstrated
Adversarial examples are a real and unsolved property of trained models. Almost every famous demonstration of one attacking a deployed system is weaker than the headline it produced, and the gap between those two facts is where defensive practice goes wrong.
Read More » -
AI Security
Gradient-Based Attacks: Which Gradient, and Where It Stops Working
In the continuous case the gradient hands you the attack. Everywhere else it only ranks candidates, and something slower does the actual work. That gap explains why these methods dominate image benchmarks and underperform in the places security teams care about.
Read More » -
AI Security
Introduction to AI-Enabled Disinformation
In recent years, the rise of artificial intelligence (AI) has revolutionized many sectors, bringing about significant advancements in various fields. However, one area where AI has presented a dual-edged sword is in information operations, specifically in the propagation of disinformation. The advent of generative AI, particularly with sophisticated models capable of creating highly realistic text, images, audio, and video, has exponentially increased the risk of…
Read More » -
AI Security
The Poisoning Tool With Millions of Downloads
Nightshade corrupts a text-to-image concept with fewer than 100 samples, and it has been downloaded millions of times by the people whose work is being scraped. The only poisoning technique with that reach inverts the whole threat model.
Read More »
