SoftBank's Son warns of superintelligence risks at Kyoto forum
The SoftBank chief called for international trust and cooperation as AI capabilities grow, warning that powerful systems could be dangerous in the wrong hands.
A warning from an AI investor
SoftBank chief Masayoshi Son said he was concerned about AI safety as systems become more capable, Bloomberg reported from an October 4 event on the sidelines of the Science and Technology in Society forum in Kyoto. Speaking alongside White House science and technology policy adviser Michael Kratsios, Son argued that countries need to build trust and cooperate on the technology. He warned that superintelligence in the wrong hands could become dangerous. Those remarks are his assessment of a potential future risk, not evidence that a particular model has caused a new security incident.
Separate from the Kyoto declaration
Son has been a prominent investor and advocate for advanced AI, making his caution notable, but Bloomberg did not report a reversal of SoftBank’s investment strategy. Kratsios was in Kyoto for a separate US-led declaration on AI-supported scientific research endorsed by 17 countries. Son’s remarks focused on safety and international cooperation; they were not part of a binding international rule or a new technical safeguard. Whether the concern leads to specific agreements, restrictions or changes in model development remains to be seen.