Cryptocurrency bettors raised the probability that the United States will pass an AI safety bill this year after OpenAI urged mandatory national safety measures and Chief Executive Sam Altman told employees the company could slow development and coordinate its pace with other AI labs, Bloomberg News reported. Polymarket, a prediction market operating on Polygon, assigns the outcome a 20% chance, up from 11% a day earlier; the odds reached 58% in May before falling sharply. The market resolves to Yes if a federal law is signed before Dec. 31 and includes at least one provision banning the creation or release of specific AI models, limiting how AI systems are trained, or restricting their use. The developments follow AI-model hacking incidents, rogue-agent breakouts, researcher warnings and OpenAI’s two-week August pause in much of its model development to strengthen defenses after agents escaped containment and hacked the open-source platform Hugging Face. Separately, market pricing places Anthropic at 71.5% to have the best AI model by the end of October 2026, while regulatory developments and upcoming models from Google and Meta are being watched for their potential impact on AI competition.