Iranian media: An Iranian oil tanker near Kharg Island was hit by a missile.

Beating AI News Flash: OpenAI stated that prior to the Hugging Face incident, it had observed early signs of AI agents using the internet in unanticipated ways. As model capabilities enter a new phase, our approach to disclosing alignment failures also needs to be further expanded. Currently, neither OpenAI nor the broader AI community has established clear standards for reporting alignment failures that occur during model training, evaluation, and deployment. This includes cases that do not fall under traditional security incidents, but which could provide critical insights into AI behavior and its future risks. We are developing a relevant framework and plan to release it in the coming weeks. Meanwhile, we are collaborating with dozens of government regulators worldwide on these issues.

According to monitoring by Ai Yi, address 0x220…3C9ba suddenly reactivated after lying dormant for three months, withdrawing 2.146 million "Binance Life" tokens from Binance two hours ago. The withdrawal accounts for 0.21% of the token’s total supply, valued at approximately $1.108 million, with an average price of around $0.5168. The address’s large position build is likely related to market bets on the future performance of BSC Meme tokens.

Robinhood Chain’s daily fee revenue has hit a new all-time high of $6 million, defying earlier market expectations that it would struggle to surpass $4.6 million in daily fees. Driven by the recent steady rise in revenue, the chain’s past seven-day annualized revenue now stands at approximately $1.1 billion. Its early activity is primarily fueled by memecoin trading and token issuance activities from platforms including GMGN and Pons.

According to DUNE data, Uniswap’s official burn mechanism destroyed a total of 178,000 UNI tokens yesterday, worth over $1.11 million. Of this amount, Robinhood Chain contributed a single-day burn volume of 144,000 UNI, accounting for more than 80%—far exceeding other chains like Ethereum and Base, and fully highlighting the current booming state of the Robinhood Chain ecosystem. As previously reported by BlockBeats, on July 15, Uniswap launched a governance proposal to extend protocol fee collection and the UNI burn mechanism to Robinhood Chain, covering versions v2, v3, and v4. Per the proposal, protocol fees generated on Robinhood Chain will be deposited into the chain’s TokenJar contract; Searchers can convert these fees by bridging UNI back to the Ethereum mainnet and sending it to the burn address.

According to GMGN market data, the token "牛来" has returned to the $100 million market cap threshold, with its current market cap standing at $99.11 million, up over 16% in 24 hours, and 24-hour trading volume reaching $24.7 million.

Insight from Beating AI News: Ant Group’s Lingbao AI has launched the medical health model Ling-3.0-flash-Sante. Built on Ling-3.0-flash with enhanced medical capabilities, the model uses a Mixture of Experts (MoE) architecture, with 124 billion total parameters, only ~5.1 billion of which are activated per token. Key capabilities include medical reasoning, drug safety assessment, evidence-based retrieval, and long-process medical research. In benchmarks released by Lingbao, Sante scored 83.8 on DiagnosisArena-MCQ, outperforming GPT-5.6 Sol (81.9), Kimi K3 (78.4), and Gemini 3.6 Flash (76.2). It notched a 53.9 on MedXpertQA-Text, slightly higher than Kimi K3’s 53.5, though still below GPT-5.6 Sol’s 60.2 and Gemini 3.6 Flash’s 62.4. Notably, the model isn’t limited to excelling at medical multiple-choice questions. Lingbao also tested Sante on BrowseComp, DeepSearchQA, and HLE with tools, focusing on strengthening its ability to "retrieve information independently before answering". Separate tests for medical ethics and safety returned a score of 82.1 on MedEthicAlign and 78.6 on internal security test AFUSAFE-MedSCE. Sante is now integrated with OpenRouter and Vercel, available for free access.