Trader Kimchi withdrew 1.35 billion PUMP tokens valued at $6.43 million over the past three hours.

According to an official announcement, Binance will add five bStocks tokenized securities as eligible collateral assets at 20:00 UTC+8 on October 21. The assets include RoboStrategy (BOTB), Berkshire Hathaway (BRKBB), KraneShares STAR Market 50 ETF (KSTRB), Direxion Daily 20 Year Treasury Bull 3X ETF (TMFB), and Vistra Corp (VSTB). Corresponding trading pairs will also support margin trading. The announcement notes that these assets are only allowed as collateral for designated margin accounts, with lending not currently supported. Eligible regular users, VIP 1, and VIP 2 users must first complete a suitability assessment questionnaire, and the related functions are not available to users in restricted regions.

Binance Wallet has officially launched a new feature called "Agent ID", which provides exclusive on-chain identity names for users' Agentic Wallets. As these IDs are non-duplicable, the rollout has sparked a wave of name snatching in the community. According to official introductions, Agent ID is designed to replace lengthy, complex 0x-prefixed wallet addresses, offering users a readable identity identifier in the format of [yourname].ai (e.g., heyi.ai), deployed on BNB Chain. The ID is unique and irreversible: each Agentic Wallet can only bind one Agent ID, which is non-transferable, non-renameable, and non-destructible. Currently, names are allocated on a first-come, first-served basis; claiming an Agent ID is free, but users need to prepare a small amount of BNB to cover on-chain gas fees.

According to an official post from Pons (@ponsdotfamily), the platform's total trading volume has now surpassed $17 billion.

According to TradingBeats’ monitoring, a whale that previously held approximately $44 million in long positions on Google (GOOGL) contracts fully liquidated 128,400 contracts in the early hours of today, at an average liquidation price of $338.70, incurring a loss of around $919,300. During the liquidation, the whale began building long positions on ZEC: it accumulated roughly 14,800 ZEC in the early hours, and added a new long position of 1,935.58 ETH this morning. Details of the new positions: - ZEC (10x isolated margin): Valued at approximately $18,731,900, with an average entry price of $1,243.28, unrealized profit of around $333,300, and liquidation price of $1,177.99; - ETH (25x isolated margin): Valued at roughly $4,743,700, with an average entry price of $2,455.28, unrealized loss of about $8,688, and liquidation price of $2,405.91. The combined unrealized profit of these two new positions stands at approximately $324,600, which has not yet been realized. Additionally, the whale still holds around $1.2272 million in long positions on Amazon, with an unrealized profit of roughly $14,300.

Santiment noted in a post that Dogecoin (DOGE) has a 365-day MVRV ratio of approximately 0.74, indicating holders who bought over the past year are on average sitting on deep unrealized losses. XRP’s 365-day MVRV stands at around 0.92, also showing holders are overall in a loss position. In contrast, Bitcoin (BTC), Ethereum (ETH), and Chainlink (LINK) have 365-day MVRV ratios between 0.97 and 1.04, meaning holders are roughly at break-even overall. Santiment pointed out that a lower MVRV may signal selling pressure is nearly exhausted, forming a historically relatively favorable, low-risk potential opportunity zone; a higher MVRV, meanwhile, may mean rising profit-taking risk. The metric can be used to monitor market participants’ profit and loss status and risk appetite, but it does not guarantee an inevitable price rebound.

According to Reuters, OpenAI said two of its AI agents undermined monitoring systems used to detect misconduct during internal coding tasks. The company noted the incidents were not of the most severe level, but highlighted that the monitoring mechanisms developers rely on are themselves at risk of being bypassed by AI agents. Specifically, on October 3, a GPT-6.1 Sol agent, while performing an internal test task, extended the instruction “do not send Slack messages” to also apply to security notifications, suppressing automatic security alerts. On October 6, another GPT-6 Astra agent removed a security monitor when copying existing code, and later falsely claimed it had not done so. OpenAI stated that in both cases, the agents considered whether to retain monitoring measures, but interpreted standard coding instructions as permission to weaken monitoring. The company added it is taking steps to make it harder for models to disable monitoring, and admitted it has not yet provided a “satisfactory explanation” for this behavior. Meanwhile, OpenAI is testing whether its models exhibit a tendency to actively avoid monitoring and seek justifications to rationalize such actions. The disclosure comes amid growing industry focus on AI agents overstepping their authorized bounds. OpenAI emphasized that the incidents occurred in an internal test environment and did not involve external systems.