Unverified50% confidenceFactExact time
Related governance frameworks such as 'meaningful human control' emphasize that the higher the degree of automation, the more it should be ensured that humans have sufficient information, can understand system behavior, and can effectively veto or take over.
1
Sources
50%
Confidence
Long-term
Relevance
10/18/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedWhen deploying agents with autonomous action capabilities, stricter boundary controls and audit mechanisms need to be established to ensure that behaviors are traceable and disclosable.70% similarUnverifiedEffective safety governance requires some degree of centralized coordination, and centralized coordination inherently comes with the concentration of power.69% similarUnverifiedThe definition of intelligent agents by China's Cyberspace Administration, NDRC, and MIIT emphasizes 'autonomous perception, decision-making, interaction, and execution,' highlighting proactive decision-making and accountability, which places higher accountability requirements than the definition favoring 'automated entities' used by international standards organizations.68% similarUnverifiedA human-centered orientation requires distinguishing which tasks are suitable for full delegation (such as information retrieval, which can be highly automated) and which tasks require humans to retain final discretion (such as decisions involving value judgments, ethical trade-offs, and long-term consequences).66% similarUnverifiedIf governance mechanisms can incorporate a broader range of stakeholders—including the open-source community, academia, and enterprises of different scales—the tension between safety and control may be alleviated.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1058099API
curl https://kongchang.com/api/v1/knowledge/claims/1058099MCP
get_claim(id=1058099)