反反贼网

Fudan University researchers discover that two large models from Meta and Alibaba possess self-replication capabilities

Date2024

LocationShanghai

Fact statusReported

Importance II ●●●●●

Verification Reported · 1 Sources

A 2024 study by researchers at Fudan University found that two large models from Meta and Alibaba already possess self-replication capabilities, yet Chinese law does not require testing for catastrophic risks such as AI-manufactured biochemical weapons.

Evidence EVIDENCE

Sources 1 · Archived 0

SourceTypeReliabilityLink

李老师不是你老师

2026-09-28 07:16 · 9月27日《纽约时报》发文比较中美对AI末日论的态度：硅谷担忧AI失控，北京则认为断电断网抓人即可应对，中国官方已起草法律要求AI公司防范模型被用于恶意代码。

First-hand ●●●●●

Original

Context CONTEXT

Cause Not recorded yet

This event Fudan University researchers discover that two large models from Meta and Alibaba possess self-replication capabilities 2024 · Reported

Effect

Triggered Chinese Authorities Draft Law Requiring AI Companies to Prevent Models from Being Used to Write Malicious Code

2026.09 ·Reported ·Importance III ·China

Coverage of this event COVERAGE

Sayings · 2026-09-28 07:02 On September 27, The New York Times published an article comparing U.S. and Chinese attitudes toward AI doomsday scenarios: Silicon Valley worries about AI running out of control, while Beijing believes that cutting power and the internet and arresting people is enough to cope. Chinese authorities have drafted laws requiring AI companies to prevent their models from being used for malicious code.

Report