AI Safety

Method prevents an AI model from being overconfident about wrong answers

What is red teaming for generative AI?

A faster, better way to prevent an AI chatbot from giving toxic responses

Students pitch transformative ideas in generative AI at MIT Ignite competition

The 2022 IBM Research annual letter

A faster way to preserve privacy online

Five ways IBM is using synthetic data to improve AI models

Explained: How to tell if artificial intelligence is working the way we want it to

A safer, lower-cost alternative to real data for pretraining computer vision models

Keeping web-browsing data safe from hackers

AI model to detect anomalies in power grid failures

Cars require regular inspection, why should AI models be any different?

The man who challenges AI every day

Chuchu Fan designs safe and reliable autonomous systems

Demystifying machine-learning systems

AI Researchers Fight Noise by Turning to Biology

Securing AI systems with adversarial robustness

Machines that see the world more like humans do

Is neuroscience the key to protecting AI from attack?