Futures
Access hundreds of perpetual contracts
TradFi
Gold
One platform for global traditional assets
Options
Hot
Trade European-style vanilla options
Unified Account
Maximize your capital efficiency
Demo Trading
Introduction to Futures Trading
Learn the basics of futures trading
Futures Events
Join events to earn rewards
Demo Trading
Use virtual funds to practice risk-free trading
Launch
CandyDrop
Collect candies to earn airdrops
Launchpool
Quick staking, earn potential new tokens
HODLer Airdrop
Hold GT and get massive airdrops for free
Pre-IPOs
Unlock full access to global stock IPOs
Alpha Points
Trade on-chain assets and earn airdrops
Futures Points
Earn futures points and claim airdrop rewards
Elon Musk Buys a $60 Billion Programming Entry Point | Rewire News Morning Report
SpaceX offers $60 billion to buy Cursor, and Florida launches a criminal investigation into ChatGPT. AI tools are simultaneously becoming strategic assets and legal targets.
1|SpaceX spends $60 billion to buy Cursor’s option; Musk completes the AI puzzle before the IPO
SpaceX has sent a $60 billion acquisition offer to Anysphere, the parent company of AI programming tool Cursor. Cursor’s annualized revenue has already reached $2 billion, making it the most profitable company in the AI programming track. The deal has not yet been finalized, but the bid itself is already a signal.
The timeline is worth noting. Three weeks ago, xAI and SpaceX completed a merger, and Musk integrated AI model capabilities into the rocket company. Now buying Cursor again is like filling in the developer entry point. Ben Thompson pointed out in Stratechery that this is not an AI acquisition, but a deal to integrate the developer ecosystem. Bloomberg also reported the same day that SpaceX’s liabilities have risen to $23 billion, and the financial pressure accumulating from three major transactions is building up. On the same day, China’s Dark Side of the Moon rushed to issue the Kimi-Coder programming model, opening it for free—signaling that competition in the AI programming market is shifting from single-tool rivalry to a fight over control of the ecosystem.
(Source: Reuters / Bloomberg / The Information / Stratechery / TechCrunch)
2|Trump opposes in the morning, delays in the afternoon: Iran ceasefire extended indefinitely, blockade unchanged
On the day the 14-day ceasefire agreement expired, Trump signed an executive order to extend the ceasefire indefinitely. But just hours earlier, he had publicly said he opposed the extension. NPR reported that the Vance team persuaded Trump to change his position within the final 3 hours, making the decision-making process unusually chaotic.
The ceasefire has been extended, but the blockade has not. The Strait of Hormuz remains closed, and the detained Iranian cargo ships have not been released. Pakistan’s four-phase mediation proposal was revealed for the first time, and its core logic is to lift the blockade in phases in exchange for Iran accepting verification. WTI crude oil broke through $95, and the market’s reaction to the ceasefire extension was not relief—it confirmed the blockade will be prolonged. On the same day, the IMF lowered its 2026 forecast for global GDP growth to 2.4%, the lowest since 2020, with energy shocks as the primary factor. A ceasefire is a pause button, not a solution. (Continued from yesterday’s report)
(Source: Axios / NPR / Al Jazeera / CNN / Fortune / IMF)
3|Meta starts recording employees’ mouse and keyboard operations to train AI agents
Reuters has an exclusive report that Meta has internally launched the Model Capability Initiative, requiring some employees to enable screen recording and input logs while working, with the data used to train AI agents that can execute internal workflows. Bloomberg followed up by saying employees privately refer to the project as “training their own replacements.”
CNN’s follow-up reporting shows that participating employees were told “this is voluntary,” but multiple employees said refusing to participate could affect performance evaluations. After Meta laid off 4,000 employees last quarter, the timing of this project has made the internal atmosphere even more tense. The issue is not whether Meta can replace employees with AI; the issue is that it is building replacement solutions using employees’ own work data—and it still needs employees to cooperate.
(Source: Reuters / Bloomberg / CNN)
4|Florida launches a criminal investigation into ChatGPT; AI responsibility first enters criminal court
Florida’s Attorney General Ashley Moody issued a subpoena to OpenAI, demanding that it turn over, by May 1, user conversation records related to the FSU shooting case. Fox 13 Tampa reported that law enforcement found ChatGPT conversations on the suspect’s phone, including content involving suggestions for weapon modification. This is the first time in the U.S. that AI chat logs have been used as part of an evidentiary chain for a criminal case.
The legal framework of the investigation is key. Moody is not invoking product liability law, but the “aiding and abetting” provisions in Florida’s criminal statutes. If this path succeeds, what AI companies face will no longer be civil damages, but criminal charges. CBS News pointed out that OpenAI’s content moderation logs will become core evidence, and the question becomes at what moment an AI model shifts from “providing information” to “assisting a crime.”
(Source: Reuters / Fox 13 Tampa / CBS News)
5|Mythos, five faces in one day: fixing bugs, becoming a target, going to the Pentagon—one model plays five roles at once
Anthropic’s Mythos model appeared in five completely different contexts within 24 hours. Mozilla used it to fix 271 Firefox vulnerabilities—Mythos’s first large-scale public use case as a developer tool. In a Stanford speech, Altman specifically named Anthropic as “selling products through fear.” The Information reported that CISA (the U.S. federal cybersecurity agency) has been locked out of Mythos’s security assessment and cannot audit this most controversial model. Bloomberg said Trump is considering allowing the Pentagon to purchase Mythos. VentureBeat’s independent benchmark testing showed that Mythos surpasses GPT-5 and Gemini Ultra in financial modeling.
The combined picture of the five pieces of news is this: an AI model is simultaneously a productivity tool, a marketing weapon, a regulatory blind spot, a military option, and an industry benchmark. No software product, within this time window after release, has generated controversy across so many dimensions at once.
(Source: Bloomberg / Axios / The Information / VentureBeat / Reuters)
Also worth knowing ↓
Adobe announces a $25 billion stock buyback, and the stock is down 30% this year. Defensive moves after the AI narrative fails. Microsoft is also increasing buybacks at the same time—large software companies are using returning capital to hedge valuation pressure during the AI transition. (Source: Reuters / CNBC)
New York’s Attorney General sues Coinbase and Gemini, alleging that prediction markets constitute illegal gambling. Citing the Martin Act, New York’s strongest financial regulation law. Polymarket’s legality faces its first state-level judicial challenge. (Source: CoinDesk / NY AG)
Eighty percent of Japan’s institutional investors plan to allocate to crypto assets within three years. A Nomura survey covers 300 institutions. Japan is shifting from being led by retail investors to seeing institutional participation, and the weight structure of Asia’s crypto market is changing. (Source: CoinDesk)
China’s Dark Side of the Moon rushes out the Kimi-Coder programming model and opens it for free. On the same day as the SpaceX news about acquiring Cursor, China’s AI programming tool track accelerates its race to secure positions. Price wars and ecosystem battles start on the same day. (Source: TechCrunch / 36Kr / Quantum Bit)
Core Scientific issues $3.3 billion in convertible bonds, shifting comprehensively from Bitcoin mining to AI data centers. This is the largest single financing for the transition of mining infrastructure into computing power infrastructure. The signal is that mining-rig companies believe the profit margins of AI compute have already surpassed those of mining. (Source: Fortune)
JPMorgan raises its S&P 500 year-end target to 7,600, with AI capital expenditure as the core driver. Wall Street’s largest bank for the first time lists AI capex as the primary reason for the index target upgrade, rather than traditional consumer or employment data. (Source: CNBC)
Click to learn about Rhythm BlockBeats job openings
Welcome to join Rhythm BlockBeats’ official community:
Telegram Subscription Group: https://t.me/theblockbeats
Telegram Discussion Group: https://t.me/BlockBeats_App
Twitter Official Account: https://twitter.com/BlockBeatsAsia