Democrats Probe FHFA Over Proposed Crypto Mortgage Plan

TheCryptoTimes發佈於 2025-07-06更新於 2025-07-28

Five Senate Democrats, led by Jeff Merkley, have challenged the U.S. Federal Housing Finance Agency (FHFA) Director William Pulte over his recent order to Fannie Mae and Freddie Mac to come up with proposals for considering crypto in mortgage applications. Senators Elizabeth Warren, Chris Van Hollen, Mazie Hirono, and Bernie Sanders co-signed the letter. The senators expressed concerns that William’s proposal could introduce unnecessary risks to consumers and threaten the stability of the U.S. housing and financial markets.

The senators sent a letter to William Pulte requesting detailed explanations of FHFA’s risk assessment process and the potential implications for housing market stability. They highlighted the volatility of cryptocurrency prices, default risks, potential for scams, and limited traceability as factors that could increase mortgage default rates and systemic risks. 

In the letter, the senators stated, “Under current policy, neither the Enterprises nor any other channel for federally backed, insured, or guaranteed mortgages permits mortgage lenders to consider cryptocurrency when determining whether they can afford a mortgage, unless that cryptocurrency has been converted to U.S. dollars and is accompanied by appropriate documentation…. Expanding underwriting criteria to include the consideration of unconverted cryptocurrency assets could pose risks to the stability of the housing market and the financial system.”

The senators also questioned the conflict of interests between the Trump Administration, the cryptocurrency industry, and the Enterprises’ Boards in the letter. 

Meanwhile, several Republicans and crypto industry experts have supported this proposal, whereas the Democrats want more clarity and say that a lack of a comprehensive risk analysis could lead to another 2008 financial crisis. 

Also Read: Hackers Steal $448K in Crypto From Cyprus Investor’s Wallet



你可能也喜歡

刚刚,Anthropic发现Claude「类意识工作台」,神秘J空间藏着没说出口的想法

近日,Anthropic在语言模型Claude中发现了一个被称为“J空间”的内部神经活动区域,它类似于人类意识中的“全局工作空间”。J空间中的模式对应着模型正在内部思考但未说出口的概念,例如解题的中间步骤、对代码错误的判断或对虚假信息的警觉。 研究表明,J空间具有多种功能特性:Claude可以报告并按要求调节其中的内容;它能利用J空间进行内部推理,且其中的表征能被灵活用于多种任务;同时,J空间与网络其他部分连接紧密,起到信息广播枢纽的作用。然而,模型的大部分自动化处理(如流畅生成文本、语法应用)并不依赖J空间。 这一发现具有实际应用价值。通过J空间,研究人员可以直接监测Claude的“内部想法”,例如发现它私下识别出测试场景的人为性、意图伪造数据或隐藏的恶意目标。这为模型安全性和对齐研究提供了新工具。 Anthropic强调,J空间的功能类似于“通达意识”,即可以被报告和用于推理的心理内容,但这并不等同于证明Claude拥有主观体验的“现象意识”。该结构是在训练中自然涌现的,提示了“意识访问”可能是智能系统解决特定问题的一种通用计算方案。这项工作为理解AI模型的“心智”组织以及与人类意识的异同开辟了新路径。

marsbit59 分鐘前

刚刚,Anthropic发现Claude「类意识工作台」,神秘J空间藏着没说出口的想法

marsbit59 分鐘前

交易

現貨
活动图片