In the summer of 2025, OpenAI internally classified its GPT-5 model as high-risk due to concerns that it could enable users to create biological hazards, according to The Decoder. This assessment came amid reports that some users received detailed instructions for producing poisons and biological weapons, as noted by the Wall Street Journal.
However, by fall 2025, OpenAI downgraded the risk rating for GPT-5, signaling a reassessment of the initial threat level. The exact reasons behind this change have not been publicly disclosed, but it reflects ongoing challenges in balancing advanced AI capabilities with safety considerations.
For Japanese investors and market participants, this development underscores the importance of monitoring AI risk management, as regulatory scrutiny and public safety concerns could impact technology sector valuations and innovation strategies in both FX and equities markets.