In a way, AI models launder human responsibility and human agency through their complexity. When outputs emerge from layers of neural networks processing billions of parameters, researchers can claim ...

Scientists give AI a dose of bad traits with the aim that it will prevent the bots from going rogue. Several chatbots, like ...

The new science of “emergent misalignment” explores how PG-13 training data — insecure code, superstitious numbers or even ...

Researchers from the University of Oxford, EleutherAI, and the UK AI Security Institute have reported a major advance in ...

But two new papers from the AI company Anthropic, both published on the preprint server arXiv, provide new insight into how ...

It's August, which means Hot Science Summer is two-thirds over. This week, NASA released an exceptionally pretty photo of ...

A new study reveals that AI models can secretly pass harmful traits to one another raising concerns about hidden risks in ...

Our weekly roundup of the latest science in the news, as well as a few fascinating articles to keep you entertained over the ...

AI is supposed to be helpful, honest, and most importantly, harmless, but we've seen plenty of evidence that its behavior can ...

Researchers are trying to “vaccinate” artificial intelligence systems against developing harmful personality traits.

A new study from Anthropic introduces "persona vectors," a technique for developers to monitor, predict and control unwanted LLM behaviors.

Google's former chief business officer, Mo Gawdat, warned that AI could soon replace white-collar jobs, including CEOs.