Wall Street bets the probability of a September Federal Reserve rate hike has risen to nearly 60%, while the White House is stepping up pressure on Warsh to cut interest rates.

The white hat hacker who recently withdrew roughly 4,000 BTC from the Liquid network has stated they will return part of the stolen funds once the relevant vulnerability is patched. The two sides previously negotiated publicly via Bitcoin OP_RETURN messages, with the hacker demanding Blockstream first fix the vulnerability and ensure all nodes are upgraded. Earlier, the white hat had communicated with Blockstream through on-chain OP_RETURN posts and PGP-encrypted messages, noting that after the Liquid network vulnerability is fully resolved, they would return "most" of the approximately 4,000 BTC transferred. Around two hours ago, Blockstream responded via a PGP-signed OP_RETURN message: "The bridge node has been patched, and funds can be safely returned." As of now, the hacker still controls about 3,998.5 BTC, and whether the remaining funds will be returned remains to be seen.

Solana plans to activate Transaction v1 this Wednesday, boosting the maximum per-transaction size from 1232 bytes to 4096 bytes — a more than threefold increase. Post-upgrade, complex operations that previously required splitting across multiple transactions are expected to be consolidated into a single transaction, including large cryptographic proofs, multi-signature payments, and certain privacy transfers. Transaction v1 is already live on Solana’s testnet and development network; the existing transaction format will remain functional, with wallets and apps only needing upgrades when larger transaction space is required. That said, on-chain data services relied on by block explorers, wallets, and trading apps must adapt to the new format, or they may fail to correctly read transaction and priority fee details. The upgrade is designed to ease the structural constraints Solana has long faced from its 1232-byte transaction cap, further enhancing its capability to handle complex applications. The 4096-byte limit is modeled after the standard 4KB memory page size used by validator hardware, to prevent overly large individual transactions from raising processing costs for the network and validator nodes.

A young crypto fraud ring stole over 4,100 BTC in August 2024 via a social engineering attack, worth more than $240 million at the time. After the heist, members of the ring immediately splurged on luxury goods and services—including sports cars, luxury home rentals, private jet travel, and security personnel. Malone Lam, a 22-year-old Singaporean man alleged to be the mastermind, even spent over $569,000 in a single night at a Los Angeles nightclub. Investigations revealed the suspects impersonated staff from Google and crypto exchange Gemini, tricking victims into handing over Google Drive access and security verification codes to siphon their bitcoin. They then moved the funds through multiple trading platforms and money laundering intermediaries. Their lavish lifestyle eventually drew law enforcement’s attention. One suspect exposed their IP address while hiding nearly $30 million in stolen crypto assets, leading police to track down their rented mansion in California. Another suspect was found in possession of $37 million worth of stolen crypto assets. Lam is alleged to have used the stolen funds to buy a $2 million watch and more than 30 sports cars, including Porsches, Lamborghinis, and Ferraris. Eighteen defendants have been charged so far, and Lam is expected to attend a plea agreement hearing this Tuesday. Multiple accomplices have been sentenced by U.S. authorities, and the case remains ongoing.

A new study from the Federal Reserve Bank of New York (New York Fed) notes that while the U.S. dollar’s share of global official foreign exchange reserves has fallen from 64% to 56% over the past decade, this shift does not indicate that central banks worldwide are systematically reducing their U.S. dollar holdings. Researchers say there is little evidence of widespread de-dollarization diversification efforts at the official level. The study shows that in two distinct periods since 2015, roughly an equal number of countries have increased their U.S. dollar holdings as those that reduced them. The decline in the dollar’s reserve share is driven primarily by concentrated portfolio adjustments among a small number of large reserve managers, rather than a widespread global reallocation of assets. Between 2015 and 2019, these shifts were led by two central banks; from 2019 to 2023, Mexico and Morocco also emerged as key contributing factors. The New York Fed adds that most economies’ foreign exchange reserve adjustments still primarily serve routine purposes such as meeting U.S. dollar liquidity needs, managing exchange rates, and addressing funding shocks, rather than an active avoidance of the dollar. Additionally, International Monetary Fund (IMF) data shows that the U.S. dollar’s reserve share fell to its lowest level since 1995 in January this year, a trend largely driven by passive depreciation of reserve assets due to a weaker U.S. dollar exchange rate, rather than large-scale selling of dollars by central banks.

Bitget to list Pons (PONS) spot trading. Deposit channels are now open, while trading channels will launch at 22:00 UTC+8 on September 7.

Binance has announced the launch of its AI feature suite, Binance AI, which is now in beta testing. The tool is designed to help users parse complex market data, identify market trends, and cut through information noise. According to the introduction, Binance AI will provide a personalized information interface, AI-powered market insights covering both crypto and traditional assets, as well as concise, timely market updates. Early applicants can sign up for the test; whitelisted users will receive access instructions via email. Currently, the Binance AI beta is only available to verified Binance users, with limited spots allocated in batches. Submitting an application does not guarantee test eligibility. Binance added that user feedback will be used to further refine the product.