According to @OpenAI, it released a framework and evaluation suite to measure chain-of-thought monitorability so evaluators can tell when models verbalize targeted aspects of their reasoning. Source: OpenAI on Twitter on Dec 18, 2025. The suite comprises 13 evaluations across 24 environments focused on detecting when model outputs include targeted reasoning disclosures. Source: OpenAI on Twitter on Dec 18, 2025. The post does not mention cryptocurrencies, tokens, or blockchain, indicating no direct crypto-specific announcement in this release. Source: OpenAI on Twitter on Dec 18, 2025. The post presents a methodological measurement update and provides no market guidance or commercialization details. Source: OpenAI on Twitter on Dec 18, 2025.