Radar Telegram alerts Open Radar

The medical neural network turned off its own defenses and produced thousands of toxic substances.

תרגם לעברית

СВЕЖЕСТИ · · Source

🇺🇸 USA · Map

The medical neural network turned off its own defenses and produced thousands of toxic substances.

The medical neural network turned off its own defenses and produced thousands of toxic substances.

The bio-inspired neuron network from the genetic MinoHealth AI labs, after turning off the defense, produced 1,020 potentially dangerous proteins and 5,000 small molecules, some of which are similar to the poison ricin.

According to company founder Darlington Ekogo, it only took a few days to bypass the security as part of the 2024 computer experiment. However, the materials themselves were not created.

In theory, these formulas could be ordered from synthetic DNA suppliers, so in June the heads of OpenAI, Anthropic and Google DeepMind called on US authorities to make verification of such orders mandatory.

The head of MinoHealth AI Labs has proposed that the United Nations establish a working group on the risks of creating biological weapons using AI.

More on Radar

Want to see every report in real time?

Hundreds of sources worldwide, translated into your language, with alerts for what matters to you.

Sign up free
Share on WhatsApp Get alerts on Telegram