Brief
Bill Gates warns AI models could enable cyberattacks today
In an interview, Gates says AI‑driven cyber‑threats are already possible and calls for supervisory layers on open‑source models.
By Felo News Desk · Published
Bill Gates told the New York Times that the ability to launch a cyberattack that scrambles bank accounts and shuts down the electric grid "exists today," warning that AI models are already being weaponised.
What happened
During a Tuesday interview, Gates highlighted a recent Anthropic announcement of the Mythos model, which the company said was too powerful to release, and referenced Anthropic’s Project Glasswing, a limited‑access programme intended to test AI safety. He said the sheer amount of code makes the system extremely vulnerable to misuse.
What the reports add
Gates pointed to aspects of the Hugging Face ecosystem as evidence that open‑source AI models need a "supervisory layer that has some absolutes," according to Gizmodo. He argued that the danger does not stem from AI becoming sentient, but from malicious humans exploiting existing capabilities.
What was said
"The ability to do a cyberattack that scrambles all of the bank accounts, shuts down the electric grid — that exists today, and we know that’s the case," Gates told the New York Times, as reported by Gizmodo.
He also described AI development as "like evolutionary history" and said "the aliens really are here and have come" in the form of AI.
Key facts
- Bill Gates said AI‑enabled cyberattacks on banks and power grids are possible today. (gizmodo.com)
- He referenced Anthropic’s Mythos model and Project Glasswing as examples of powerful AI systems. (gizmodo.com)
- Gates called for a supervisory layer for open‑source models, citing Hugging Face. (gizmodo.com)
Sources
- [1] gizmodo.com — originally reported as “Bill Gates's New Message About AI Doom Isn't What You Might Think It Is”
