OpenAI calls for international standards on AI that could improve itself
OpenAI urges US-led international standards for recursive self-improving AI, including oversight, measurement, and incident reporting.
OpenAI is calling for international technical standards governing advanced AI that could recursively improve itself. The company says fully autonomous recursive self-improvement is not happening today and should be pursued only when it can be done safely, warning that weak safeguards could make systems less aligned and more dangerous. It wants the United States to lead shared measurement methods, incident reporting, and human-oversight rules, building on national AI safety institutes, CAISI, and ISO. A recent UN scientific panel raised similar loss-of-control concerns and proposed cooperation modeled on aviation, nuclear energy, and cybersecurity.
- OpenAI says fully autonomous recursive self-improvement is not happening today.
- It proposes US-led standards built with safety institutes, CAISI, and ISO.
- Suggested controls include measurement, incident reporting, and human oversight.
- A UN panel separately warned about loss of control over agent swarms.
Full article196 words · extracted from the-decoder.com · click to collapse
OpenAI is pushing for international standards on advanced AI development, focusing on recursive self-improvement (RSI). With RSI, AI systems increasingly automate the research needed to build the next generation of AI, potentially accelerating progress dramatically. "Fully autonomous RSI is not happening today," OpenAI writes, and it should only be pursued "until it can be done safely." Without mature safeguards, humans could lose track of AI development, and AI could become "more dangerous, less aligned, and, on the whole, a danger to people."
OpenAI proposes that the US lead on global technical standards, building on national AI safety institutes and organizations like CAISI and ISO. That would include shared measurement methods, incident reporting, and rules for human oversight of automated AI research. A UN scientific panel recently raised similar concerns, warning about loss of control over autonomous agent swarms and proposing international AI safety cooperation modeled after aviation, nuclear energy, and cybersecurity frameworks.
AI News Without the Hype – Curated by Humans
Subscribe to THE DECODER for ad-free reading, a weekly AI newsletter, our exclusive "AI Radar" frontier report six times a year, full archive access, and access to our comment section.
Text extracted automatically; images, tables and formatting may be missing. Original: https://the-decoder.com/openai-calls-for-international-standards-on-ai-that-could-improve-itself/