10 #Bitcoin ETFs inflows 7,066 $BTC(+$703.6M) and 9 #Ethereum ETFs inflows 14,425 $ETH(+$48.32M) yesterday.

According to on-chain analyst Ai Yi (@ai_9684xtpa), address 0x5ce…1a723 added more BTC long positions today, pushing its BTC long position size to the third largest on Hyperliquid. · Holds a 40x long position of 1,169.52 BTC, with a position value of approximately $98.15 million, average entry price of $83,822.9, and current unrealized profit of around $182,000; · Also holds a 10x long position of 12,673.27 ZEC, with a position value of roughly $19.69 million, average entry price of $1,217.84, and current unrealized profit of approximately $4.231 million.

NFT marketplace Magic Eden has reportedly experienced an NFT security vulnerability, as a white hat hacker transferred 3832 NFTs from hundreds of wallets. Yuga Labs' Vice President of Blockchain, Quit, stated that the transfer is a white hat operation, noting the relevant NFTs are currently stored in an address starting with "0x71cF" and the assets remain secure. All NFTs will be fully returned once the risk is resolved.

Bitget CEO Gracy Chen stated in a post that the crypto exchange is collaborating with independent third-party experts Mandiant and SlowMist to launch a full investigation into the recent platform security incident. The company will continue to share the latest updates via official channels, stressing that protecting users is Bitget’s top priority at present. User balances remain intact, and the Bitget User Protection Fund will cover losses stemming from this platform-level event. Bitget Wallet, a self-custody wallet, operates on infrastructure fully separate and independent from Bitget Exchange, so it has not been impacted by the incident. As of now, deposit, trading, and reward functions on Bitget Exchange are still running normally. The withdrawal function remains suspended amid additional security checks and will be reopened as soon as the platform confirms it can safely resume operations.

Beating AI News Flash: Anthropic Resumes Billing for Three Categories of "Zero-Output Safety Rejections" Claude will now charge at its standard model rate even when it rejects requests before generating any output, if the rejection stems from biosecurity, frontier large language model (LLM) development, or "reasoning extraction"—a category referring specifically to attempts to elicit Claude’s internal reasoning processes. On June 2 this year, Anthropic had scrapped all charges for zero-output rejections, but is now reinstating billing exclusively for these three categories. The change only applies to these cases: other requests (such as those for cyberattacks or general harmful content) that get rejected before output remain free of charge. If Claude has already generated partial output before being blocked, it will always be billed normally. Anthropic noted that coordinated attacks targeting its system have emerged in recent weeks, so billing is being deployed as an additional defense measure. Official tests found that 99.7% of Claude Code, Claude.ai, and Cowork accounts did not encounter these three billing-triggering rejections, and the false positive rate for related safety classifiers is below 0.1%. However, false positive reports have already appeared on GitHub. One Claude Code user documented 23 false "reasoning extraction" judgments, including for routine tasks like podcast scripts and budget spreadsheets, with 7 cases terminating the current task. Another user reported that general software development tasks and Claude Code’s built-in workflow instructions also triggered similar rejections.

Saturn Foundation, an on-chain digital credit protocol, announced that the STRN Token Generation Event (TGE) is scheduled for the fourth quarter of 2026, with the second season of its program reserving up to 5% of the total token supply for distribution. If the TGE occurs before December 8, 2026, the second season will conclude shortly prior to the TGE, at which point a snapshot will be taken. Should the second season end early, its TGE allocation will be reduced proportionally based on the program’s actual duration; unallocated amounts will be returned to the Saturn ecosystem allocation pool for future seasons. For example, if the program completes 75% of its original cycle, only 75% of the reserved allocation will be distributed. The third season will launch immediately after the second season’s snapshot is finalized; eligible positions will continue to accrue second season points until the snapshot is taken, and users are not required to take any action at this time. Specific details regarding the token economic model, distribution, and claim arrangements will be disclosed in subsequent announcements.

Beating AI News: Elon Musk has outlined a timeline for SpaceX AI to catch up with OpenAI and Anthropic. He said the firm is on track to build models matching the level of Anthropic’s Fable and OpenAI’s GPT-6 in the next 2 to 3 months; if progress continues accelerating, SpaceX AI could “take the lead” in roughly 6 months. Musk also revealed the latest GPU counts for Colossus: Colossus 1 has 150,000 H100s, 50,000 H200s, and 30,000 GB200s; Colossus 2 has 110,000 GB200s and 440,000 GB300s. The two data centers hold a combined ~780,000 GPUs, more than the prior external estimate of 670,000. Another 220,000 GB300s will go fully operational next week, bringing total GPUs to exactly ~1 million. An additional 220,000 will be added in November; if all goes well, a further 220,000 could be added by end-December, pushing the total to ~1.44 million.