DeepFlowTech

Author Matt Brown points out that AI has reduced code costs, but simultaneously made core financial service assets like banking licenses, underwriting data, and genuine transaction volume more valuable. Fintech companies need to reshape their thinking, focusing on risk-taking, regulatory relationships, and proprietary data to build stronger moats, rather than relying solely on cheap code.

Author: Nishil Jain
Compiled by: TechFlow
TechFlow Editor's Note: DeFi on-chain lending reached a historic high of $90 billion in Q4 2025, yet institutional capital represents only 11.5% of TVL—this contrast reveals the core of this article. Regulatory barriers are gradually crumbling (GENIUS Act passage, SEC withdrawal of multiple investigations), but what truly blocks institutions is the lack of risk isolation infrastructure: no fixed rates, no risk stratification, no tools that can be embedded into internal compliance frameworks. The author systematically outlines how Aave V4, Morpho's curator model, Pendle's yield splitting, and Maple's structured credit each fill this gap—one of the most comprehensive institutional DeFi roadmaps to date.
Full text below:
According to

Tencent's "Lobster Task Force" recently responded to questions about its AI Agent product, launching multiple products and planning to expand the series. Despite current limitations in technical capabilities, "Lobster" still provides non-technical users with a low-barrier AI experience. Offline activities have successfully attracted diverse audiences, demonstrating the product's popularity.

Chicago-based crypto trading and lending firm BlockFills filed for Chapter 11 bankruptcy with the U.S. Bankruptcy Court for the District of Delaware on March 15, 2026, with estimated assets between $50 million and $100 million and liabilities between $100 million and $500 million. The company suspended customer deposits and withdrawals due to poor market and financial conditions, and plans to conduct bankruptcy restructuring under court supervision to protect customer interests.

The article discusses the structural risks the United States faces in the robotics revolution, emphasizing China's control over rare earth materials and the robotics hardware supply chain. China accounts for 70% of global rare earth mining and a high proportion of refining and manufacturing capabilities, making it difficult for the United States to independently produce the materials needed for humanoid robots. To ensure the continued development of the robotics industry, the United States must invest in rebuilding the rare earth supply chain and focus on key enterprises in related sectors.

A wallet address is suspected of being involved in a liquidation event on Venus Protocol. The attacker borrowed stablecoins using ETH, then accumulated large amounts of THE and drove up the price, causing the price to collapse and triggering a cascade of liquidations, resulting in $2.15 million in bad debt. The attacker profited approximately $5.07 million.

Aave released a report stating that a user lost approximately $50.43 million when exchanging aEthUSDT for aEthAAVE, as the order size exceeded market liquidity. Although the Aave protocol was unaffected, Aave will refund the user's $110,000 in transaction fees and launch an "Aave Shield" feature to prevent similar incidents from occurring.