历史小径·世界史英语30篇（1）

23 / 30

人工智能协议会谈：当算法遇见外交

In 2023, the EU, U.S., Japan, and Canada launched the Hiroshima AI Process to set shared guardrails for powerful systems.
Governments debated whether to regulate AI by risk level, banning only high-stakes uses like autonomous weapons or biometric surveillance.
China published its own ethical guidelines stressing 'social stability' and state oversight over innovation speed.
UNESCO’s 2021 AI Ethics Recommendation became the first global framework, though it lacks binding enforcement powers.
At the OECD, experts argued that transparency rules must apply equally to Silicon Valley startups and Beijing-based labs.
Developing nations warned that strict AI laws might widen the digital divide unless training data and compute access improved.
The Council of Europe drafted a treaty on AI human rights impact, requiring audits before public-sector deployment.
Tech firms joined talks reluctantly, fearing fragmented rules that would complicate global product launches and updates.
Unlike climate or nuclear treaties, AI governance remains voluntary, relying on trust rather than inspections or sanctions.
These early talks show how digital tools force diplomats to rethink sovereignty, accountability, and what ‘international order’ means today.