English

Bodies with power

Beijing Institute of AI Safety and Governance (Beijing-AISI)

Type
Government body
Place
China — People's Republic of China (Beijing municipality)
Local name
北京前瞻人工智能安全与治理研究院
Last checked
2026-09-22
Next check due
2027-03-21

Ways to reach them

  • Institute website (Chinese domain) · Web form
    Works in: zh, en
    Verification expired · 2026-09-22

    This route is retained as reference only. Its verification is missing, more than 180 days old, or the contact window is not currently open; do not rely on it until it is re-checked.

    general

    VERIFIED exists and serves an English locale (?lng=en). JavaScript-rendered — no contact details were machine-readable.

  • Institute website (English domain) · Web form
    Works in: en
    Verification expired · 2026-09-22

    This route is retained as reference only. Its verification is missing, more than 180 days old, or the contact window is not currently open; do not rely on it until it is re-checked.

    general

    VERIFIED exists. Also JS-rendered; no contact details extractable.

  • GitHub organisation · Web form
    Works in: en
    Unverified

    technical/research

    UNVERIFIED (search result). For a researcher this is likely the most practical contact surface — GitHub issues on their published benchmarks.

  • Homepage · Homepage
    Unchecked

What it does

Research institute founded in the Beijing AI ecosystem, associated with Yi Zeng (Institute of Automation, Chinese Academy of Sciences), who is also a member of the UN High-Level Advisory Body on AI and a leading Chinese voice on AI risk. Works on AI safety evaluation and benchmarking (it hosts a 'foresight safety bench'), governance research, and international engagement. Research body with no regulatory power. A separate Beijing Key Laboratory of AI Safety and Super Alignment also exists at Beijing Normal University — do not conflate them.

Its powers

research

Honest assessment

Academically open in the normal way — its people publish, attend international venues, and are contactable as individual researchers. Institutionally it publishes no contact channel I could verify. The realistic route for a foreigner is academic: cite or engage with their published work, or contact Yi Zeng directly through his CAS affiliation. Not a channel for public complaints.

Notes

The '.institute' vanity domain and the shared redirect from the CnAISDA network domain suggest a deliberate outward-facing internationalisation effort by this group.

Sources

Something wrong here?