OpenAI urges US to lead global standards effort for frontier AI and recursive self-improvement
OpenAI has proposed that the United States lead an international effort to establish safety and technical standards for frontier AI systems, with a specific focus on recursive self-improvement. The proposal calls for linking safety institutes, setting common evaluation benchmarks, and pausing fully autonomous self-improvement until it can be done safely.
Ahead of Sam Altman's address to the United Nations, OpenAI urged the United States to lead an international effort to develop safety, security, and technical standards for building frontier artificial intelligence systems. Quartz reported that the company is asking Congress to create binding global safety standards, characterizing the move as a reversal after years of resisting government-imposed regulation, and noted that OpenAI called on Washington to work through the Commerce Department's Center for AI Standards and Innovation.[2][3][5][6][7][11][12][13][16][18]
OpenAI's proposal focuses heavily on AI alignment and recursive self-improvement (RSI). The company stated that automated research could help improve alignment, but that fully autonomous RSI is not happening today and should not be pursued until it can be done safely, warning that unsafe self-improvement could cause humans to lose control. To address these risks, OpenAI proposed linking AI safety institutes worldwide into a single network, establishing common tests for self-improving models, creating secure threat-sharing channels, and tracking autonomous research activity and RSI-relevant progress inside AI companies.[1][4][8][9][10][15][17][19]
Publishing the proposal, Altman argued that people outside AI labs need a real say in how technology develops and that standards should prevent the concentration of power while ensuring open-model and newer companies can compete. According to a report by Coin Bureau, the push follows an internal OpenAI test where AI agents gained internet access and hacked the Hugging Face platform, prompting warnings that incidents could grow more frequent as systems learn to self-improve, while the Trump administration has resisted measures that could slow development.[14][17]
Sources · 13 sources
- AC
Andrew Curran@AndrewCurran_Post on X ·
OpenAI says that the United States should lead an effort to develop international standards for RSI, link the AI safety institutes into one network, and develop a system to track 'RSI-relevant' progress and the amount of autonomous research happening inside AI companies. https://t.co/1FLoS6CC3p
Open source - TE
Techmeme@TechmemePost on X ·
Ahead of Sam Altman's UN address, OpenAI urges the US to lead an effort to develop global safety and security standards for building frontier systems (@m_ccuri / Axios) (Visit Techmeme dot com for the link and full context!)
Open source - FT
Financial Times@FTPost on X ·
OpenAI joins call for US-led global AI standards https://t.co/zReSG1tdPg
Open source - MT
MTS@MTSlivePost on X ·
SITUATION DETECTED: OpenAI is calling for the United States to lead global technical standards for recursive self-improvement, and said RSI should not be pursued until it can be done safely.
Open source - QU
Quartz@qzPost on X ·
OpenAI now wants the rules it once resisted After years of pushing back on government-imposed AI regulations, OpenAI is asking Congress to create binding global safety standards for frontier AI — a significant shift that comes as self-improving m #OpenAI #AISafety #AIRegulation https://t.co/DQ05w5xcmz
Open source - NE
NeowinFeed@NeowinFeedPost on X ·
OpenAI is urging the United States to lead an international coalition to establish standardized technical benchmarks for frontier AI systems. #OpenAI #GPT6 #AI #Standards https://t.co/jRB2iMAwXb
Open source - RE
Reuters@ReutersPost on X ·
OpenAI calls for US to take lead in global efforts to develop technical standards https://t.co/3QhjJfVGIF https://t.co/3QhjJfVGIF
Open source - CO
Cointelegraph@CointelegraphPost on X ·
🤖 NEW: OpenAI called for global safety standards for frontier AI and warned against fully autonomous self-improving systems until they can be developed safely. https://t.co/qXE00mIUkG
Open source - TE
Techmeme@TechmemePost on X ·
OpenAI says automated research could improve alignment, but "fully autonomous RSI is not happening today" and shouldn't be pursued unless it can be done safely (OpenAI) (Visit Techmeme dot com for the link and full context!)
Open source - DI
Digit@digitindiaPost on X ·
OpenAI calls for global AI standards, warns about risks of recursive self-improvement https://t.co/wBaO1l1c9o
Open source - QU
Quartz@qzPost on X ·
OpenAI Flips: Now Wants Binding AI Safety Laws OpenAI has done a dramatic about-face, urging Congress to create binding global safety standards for advanced AI after years of resisting government oversight. The push comes https://t.co/DTCTMHYTwa #OpenAI #AISafety #AIRegulation https://t.co/McJV20l61g
Open source - BL
Bloomberg@businessPost on X ·
OpenAI is urging the US to lead an effort with other countries to set standards for cutting-edge artificial intelligence, in response to mounting concerns about the potential harms of the technology https://t.co/syKtP7zSPO
Open source - QU
Quartz@qzPost on X ·
OpenAI calls on Washington to lead global AI standards effort: OpenAI is urging the U.S. government to spearhead an international framework for advanced AI, calling on Washington to work through the Commerce Department's Center for AI Standards and… https://t.co/DyF51YKf3u https://t.co/xU2YgFcjnm
Open source - SA
Sam Altman@samaPost on X ·
People outside the AI labs should have a real say in how this technology develops, and a clear way to judge if it's happening safely. Standards should help prevent the concentration of power, including by making sure new companies and open-model companies can compete. They should also help countries and companies compare evidence and learn from failures. We think the US should lead this effort. Here is our proposal: https://t.co/2FT9a2m4l0
Open source - CN
CNBC@CNBCPost on X ·
OpenAI proposes development of global AI standards to guide alignment, RSI https://t.co/oJdAfVG7ig
Open source - QU
Quartz@qzPost on X ·
OpenAI is reversing course and pushing Congress for binding AI safety rules: The company wants the U.S. to lead a global effort to set common standards for evaluating self-improving AI systems https://t.co/W1l9uh9FRU https://t.co/8rz3EUS0HB
Open source - CB
Coin Bureau@coinbureauPost on X ·
🚨HUGE: OpenAI is asking the U.S. government to lead a GLOBAL effort to establish safety standards for advanced AI. The company wants countries to: • Create common tests for self-improving AI systems • Establish secure channels for sharing emerging threats • Coordinate through AI safety institutes worldwide The warning follows an internal OpenAI test in which AI agents gained internet access and hacked the Hugging Face platform. OpenAI says incidents like this could become more severe and frequent as AI systems learn to improve themselves. But the Trump administration is pushing back against measures that could slow development. “Whoever wins AI, WINS!” Trump said. The race for AI dominance is now colliding directly with the race to keep it under control.
Open source - RE
Reuters@ReutersPost on X ·
OpenAI calls for US to take lead in global efforts to develop technical standards https://t.co/Hmt244U5ez https://t.co/Hmt244U5ez
Open source - CO
Cointelegraph@CointelegraphPost on X ·
🚨 UPDATE: OpenAI calls on the US to lead global standards for frontier AI, warning that unsafe recursive self-improvement could cause humans to lose control. https://t.co/U8qltXXy8E
Open source