Closed-Source Models Are Not Safer Than Open-Source

xeophon · x · 2026-07-17

The author reshared their research blog regarding AI model safety. By investigating actual malicious use cases, the article points out that closed-source models are not safer than open-source models. They are equally susceptible to being exploited by hackers for attacks, generating disinformation, and conducting scams.

Original post →

More from Safety

Safety channel →