Tether Freezes $4.04 Million USDT, Involving 6 Tron Addresses

Insight Beating AI News Flash: Anthropic’s Claude has faced a fresh wave of mass account ban complaints recently. The South China Morning Post reports that on October 1, multiple users in Hong Kong, China suddenly received account suspension notices. They had previously accessed Claude via virtual private networks (VPNs), with some having used the service for up to half a year. A user on Chinese tech forum V2EX also noted their three-year-old account was deactivated. A key focus of the recent bans is region restrictions: Anthropic’s official supported regions do not include mainland China, Hong Kong, or Macau. Some users received emails explicitly citing "violation of supported region policy". Even switching VPN exit locations to the U.S. or Singapore does not meet compliance requirements, as the company mandates users be physically present in supported regions. However, some users have complained of suspected wrongful bans. A paying user claiming to reside in the U.S. posted on GitHub that their account was banned due to region policy. After appealing, Anthropic refused to restore the account under the pretext of "violating usage policy" without specifying the exact infraction. The user added that the appeal channel was closed, and contacting other support teams yielded no effective resolution. Anthropic has not disclosed the scale of the recent bans nor confirmed it is conducting a new round of mass cleanup. The company updated its usage policy on October 8, reaffirming the scope of its region restrictions.

Robinhood Crypto announced it is exploring a partnership with the digital assets division of T. Rowe Price, the asset management giant overseeing roughly $1.9 trillion in assets, to launch stock tokens linked to actively managed ETFs on Robinhood Chain. If the plan moves forward, this would mark Robinhood’s first such stock token. Robinhood stated that the initiative remains forward-looking and exploratory, requiring completion of legal due diligence, approval from the issuer’s board of directors, and regulatory clearance across all relevant jurisdictions. Its stock tokens are tokenized debt securities, and are currently unavailable to U.S. users, with additional restrictions applying in Canada, the U.K., Switzerland, and other regions.

Julia Leung, CEO of Hong Kong’s Securities and Futures Commission (SFC), said in a speech at the Asia Securities Industry and Financial Markets Association (ASIFMA) conference on Friday that the regulator is studying measures to reform market microstructure, including extending securities trading hours, which will first be rolled out in the derivatives market. This move will help investors respond timely to fast-changing global markets spanning different time zones, ensuring Hong Kong retains its competitive edge. In the long term, the extended trading hours policy will be supplemented by other upcoming innovative measures, such as tokenized currency and digital Hong Kong dollar (e-HKD) settlement arrangements. Together, these measures will give investors greater flexibility to adjust positions and manage risks based on the latest market information. Hong Kong Exchanges and Clearing (HKEX) will release a consultation paper on extending spot market trading hours in the fourth quarter, and welcomes all stakeholders to submit feedback actively.

According to Bitget market data, Hong Kong-listed Xiaomi Group-W (HK1810) surged over 8% to trade at HK$25.54 per share. Driving the rally: Xiaomi’s first range-extended electric vehicle, the Pengcheng (branded Sky Nomad), secured more than 70,000 confirmed orders in its first month on the market, while its September vehicle deliveries have already exceeded 40,000 units.

Beating AI News Flash: Anthropic has updated its usage policy, explicitly banning users from persistently and unreasonably abusing or cruelly treating its AI models (such as Claude) for the first time. The new rule will take effect on November 12. Anthropic stressed that the ban only applies to extreme cases of repeatedly maliciously treating models with no clear purpose. Users getting angry at or criticizing models for wrong answers, or conducting normal testing and research, will not be affected. In fact, Anthropic already allowed Claude to proactively end extremely malicious conversations as early as last August; this time, such behavior is formally classified as a violation. The company noted that the primary measure is still to have Claude end the current chat, and users can restart the conversation. As for whether additional penalties like account bans will be imposed, Anthropic has not yet specified. This update ties into Anthropic’s stance on AI consciousness: the company has been researching whether models might possess subjective experiences, but has not reached a definitive conclusion. The new rule has sparked debate on Reddit: some support it, arguing that AI should not be used as an outlet for emotional venting; others question whether Anthropic is over-personifying AI. The same policy update also clarifies bans on fake account promotion, weapons control software, and unauthorized surveillance of others.

Citi believes that if the short-term growth rate of core PCE continues to hover near the 2% annualized level, the tightening path after the September rate hike may shift. The key question is whether the Fed’s previously expected subsequent rate hikes will still materialize, rather than an immediate pivot to rate cuts. Citi economist Andrew Hollenhorst noted in a report that the U.S. economy is not overheated, adding that “we have not seen strong grounds for further rate hikes.” The September meeting minutes revealed that at least some officials also did not judge the economy as clearly overheated, but agreed to raise rates based on risk management amid upside inflation risks. This means if underlying inflation continues to cool, the preventive justification for the September rate hike could quickly fade. Most officials still expect a possible additional rate hike before year-end, but the minutes show no urgency for a consecutive October rate hike. Citi prioritizes core PCE month-on-month data over the still-high year-on-year figure: August core PCE rose 0.2% month-on-month, July was revised to 0.1%, and August year-on-year rose 3.0%. If it runs at an annualized rate of around 2% for several consecutive months, inflation would be seen as falling at the “sufficient pace” required by the Fed. Citi also remains cautious about energy shocks passing through to core inflation, arguing that companies have not yet broadly passed on energy costs to core prices. Its baseline scenario projects core inflation to stay low over the next four months, and flags downside risks to September core inflation. A so-called “dovish surprise” would likely mean the rate hike cycle ends earlier than the dot plot and market expectations, rather than a rapid shift to rate cuts. Fed Chair Jerome Powell has not yet clarified the standard for “sufficient pace”; an annualized core PCE rate of around 2% over several consecutive months can serve as a reasonable reference.