According to CoinWorld, OpenAI's account and billing system experienced a synchronization failure, leading to service level chaos and exacerbated API throttling on the ChatGPT web platform. Some Pro members paying $200 per month had their usage quotas abnormally reset to 75%, or their quotas were directly increased by four times. However, some Pro members reported their account behavior being downgraded to Plus restrictions, and they were unable to select the GPT-5.5 Pro flagship inference model in chat settings. Tibo Sottiaux, OpenAI's core product lead, stated that the accelerated consumption of user quotas was due to a misjudgment by the system's built-in anti-fraud and anti-abuse mechanisms. The official team has urgently reset the usage limits for all Codex users free of charge. The abnormal quota consumption issue began around June 16th and erupted on social media platforms on June 25th, with hundreds of developers reporting an abnormal surge in Codex token consumption, with some accounts experiencing a 10-20 times faster consumption rate than usual. The slow response from the OpenAI development team sparked numerous protests from developers on social media platforms.

MyToken Portfolio Daily Report 10% of users gained over +1.14%；50% of users gained over +0.52%；90% of users gained over +0.00%。

According to CryptoWorld, Anthropic released its latest AI Economy Index report, combining survey data from 9,700 users with telemetry data to reveal the impact of artificial intelligence on work patterns and career prospects. The report found that the higher the value of a job, the more AI computing power it consumes: high-paying positions consume an average of 2.07 times more computing power than low-paying positions. The survey showed that over one-third of users expect AI to take over most of their work within a year. Users exhibit significant differences in the level of autonomy they delegate when using different AI tools. When writing articles on a web platform, users are accustomed to revising back and forth with AI up to 13 times, while when using terminal tools like Claude Code, they typically only issue one command, allowing AI to generate the code directly in one step.

According to Coinbase.com, CEO Brian Armstrong stated that the company has nearly halved its enterprise AI spending by optimizing its infrastructure, while the use of AI tokens continues to grow exponentially. Key cost-cutting measures include making open-source models such as GLM 5.2 and Kimi 2.7 the default options for internal LLM gateways, matching tasks with the most cost-effective models through intelligent routing, and significantly improving the cache hit rate of tools like LibreChat from 5% to 60%. Armstrong emphasized that the goal of managing AI costs is not to restrict usage, but to make exponential growth sustainable by reducing waste.

According to CoinWorld, Wu has learned that Anthropic has announced that the US government has notified the company that its most robust cybersecurity model, Claude Mythos 5, can be redeployed to a number of US agencies responsible for operating and protecting critical infrastructure. Anthropic stated that it is working to restore access to these agencies as quickly as possible and will continue to cooperate with the US government to expand access to Mythos 5 while also pushing for Fable 5 to be reopened for general use.