☕️ Unauthorized users breach Anthropic's restricted Mythos AI model

Hi there, this is your daily ☕️ Techpresso.

In today's Techpresso:

🔓 Unauthorized users breach Anthropic's restricted Mythos AI model

🚀 SpaceX secures option to buy Cursor for $60 billion

🖱️ Meta tracks employee keystrokes and clicks to train AI agents

🔀 Google splits its TPU line in two for the agentic era

🎨 OpenAI unveils ChatGPT Images 2.0

⚖️ New York sues Coinbase and Gemini over prediction markets

Plus: 🎁 20 other news you might like, 🧰 6 tools, and 📚 4 papers.

🔓 Unauthorized users breach Anthropic's restricted Mythos AI model LINK

Unauthorized users from a private online forum breached Anthropic's restricted Mythos AI model through a third-party vendor environment on the same day the cybersecurity tool was publicly announced.

The group breached access by guessing Mythos's online location based on the format Anthropic used for other models, and one member had access through employment at a third-party contractor.

Anthropic said it is investigating the breach but has found no evidence the unauthorized activity impacted its systems, even as the group provided Bloomberg screenshots and a live demonstration.

🚀 SpaceX secures option to buy Cursor for $60 billion LINK

SpaceX has reached a deal with Cursor to build a new "coding and knowledge work AI," and the agreement includes an option for SpaceX to acquire the popular development platform for $60 billion.

The partnership comes after xAI began renting computing power to Cursor and two senior Cursor engineering leaders, Andrew Milich and Jason Ginsberg, left to join xAI and report directly to Musk.

Neither Cursor nor xAI has proprietary models matching Anthropic or OpenAI, and Cursor still sells access to Claude and GPT even as both companies roll out their own competing coding tools.

🖱️ Meta tracks employee keystrokes and clicks to train AI agents LINK

Meta is recording employee keystrokes, mouse clicks, and occasional screen snapshots through a tool called Model Capability Initiative to collect training data for its AI agents.

The data will teach AI models to replicate how humans interact with computers, including choosing from dropdown menus and using keyboard shortcuts, according to an internal memo.

Legal experts say this subjects white-collar workers to real-time surveillance once limited to gig workers, and European law would likely prohibit the practice under GDPR rules.

🔀 Google splits its TPU line in two for the agentic era LINK

Google announced its eighth-generation TPU as two separate chips — TPU 8t for training and TPU 8i for inference — marking the first time the company has split its TPU line across different silicon.

TPU 8t scales to 9,600 chips per pod at 121 exaflops, while TPU 8i pairs 288 GB of high-bandwidth memory with 384 MB of on-chip SRAM and cuts latency up to 5x using a new Boardfly network topology.

Broadcom reportedly designs the training chip codenamed Sunfish, MediaTek handles the inference chip codenamed Zebrafish, and TSMC fabricates both — with Intel and Marvell filling out the surrounding data-center supply chain.

🎨 OpenAI unveils ChatGPT Images 2.0 LINK

OpenAI is releasing ChatGPT Images 2.0, an update to its image-generating software that adds a reasoning mode designed to produce accurate charts, scientific diagrams, and other complex visuals aimed at professionals.

The update rolls out Tuesday through OpenAI's flagship chatbot and its Codex AI coding assistant, with improved ability to follow instructions and include more details in generated images.

ChatGPT Images 2.0 can also produce visuals that more faithfully reflect a range of styles and render text in multiple languages, according to the company's announcement.

⚖️ New York sues Coinbase and Gemini over prediction markets LINK

New York Attorney General Letitia James filed separate lawsuits against Coinbase and Gemini on Tuesday, accusing both crypto companies of running illegal gambling operations through their prediction markets without state gaming licenses.

James says the platforms let users as young as 18 bet on sports and elections, violating New York law that requires mobile sports bettors to be at least 21, and she wants profits returned and civil fines tripled.

The cases land in the middle of a three-way fight between states, the federal CFTC, and the industry, with the CFTC suing multiple states and arguing it alone has authority over prediction markets nationwide.

Other news you might like

• Microsoft removes Call of Duty from Game Pass, lowers subscription pricingLINK

• Florida probes ChatGPT role in mass shooting. OpenAI says bot "not responsible."LINK

• John Ternus says Apple is “about to change the world once again”LINK

• Mozilla: Anthropic's Mythos found 271 zero-day vulnerabilities in Firefox 150LINK

• Tron's Justin Sun sues Trump-linked World Liberty Financial over frozen assetsLINK

• Kalshi takes on Coinbase, Robinhood with new plan to offer crypto perpetual futures: The InformationLINK

• Trump Is Warming Up to Anthropic Again, Says the Company Could ‘Be of Great Use’LINK

• Sullivan & Cromwell law firm apologizes for AI 'hallucinations' in court filingLINK

• Anthropic outspends OpenAI in biggest-ever lobbying quarterLINK

• Tencent, Alibaba in Talks to Back DeepSeek at $20 Billion Valuation, Report SaysLINK

• Is Claude Code going to cost $100/month? Probably not - it's all very confusingLINK

• Ransomware negotiator pleads guilty to helping ransomware gangLINK

• Anker made its own chip to bring AI to all its productsLINK

• Exclusive: SpaceX says unproven AI space data centers may not be commercially viable, filing showsLINK

• Meta projected $16 billion in scam ad revenue. Now the lawsuits are piling up.LINK

• Deep Research Max: a step change for autonomous research agentsLINK

• Bond, a new social media platform, wants to use AI to help you kick your doomscrolling habitLINK

• New Gas-Powered Data Centers Could Emit More Greenhouse Gases Than Entire NationsLINK

🧰 Trending tools

ChatGPT Images 2.0: generates, refines, and validates images in one workflow, with flexible aspect ratios and multiple outputs per prompt for faster asset production.LINK

InstantDB: a full-stack backend that pairs with AI tools to ship apps faster, bundling auth, permissions, storage, and real-time features out of the box.LINK

Zernio Ads API: unified API that handles publishing, inbox, analytics, and account management across 14 social platforms, eliminating direct platform API integrations.LINK

Nova Recruiter: automates end-to-end talent sourcing across 800M+ profiles, using AI agents and multi-channel outreach to cut recruiting time significantly.LINK

Toki 2.0: an AI scheduling agent that converts rough ideas into structured daily plans while learning your preferences and automating calendar actions.LINK

Framework Laptop 13 Pro: a modular, user-repairable laptop that lets you swap components and upgrade hardware, reducing e-waste and long-term costs.LINK

📚 Trending papers & reports

MemFactory: unified framework for training and evaluating memory-augmented LLM agents with modular, plug-and-play memory lifecycle components.LINK

FPGA matrix multiplication accelerates a key AI language model bottleneck, achieving a 7x speedup over ARM CPU and 200x over standard NumPy, reaching 3.1 GFLOPs on a low-power edge chip.LINK

SMC speculative decoding speeds up AI text generation by 5.2x over standard methods, staying within 3% of the original model's accuracy across reasoning, instruction-following, and coding tasks.LINK

Short-form video addiction can be interrupted by showing users self-related cues like their live camera feed or a black screen, with 84 participants choosing to voluntarily stop scrolling as a result.LINK

Want to get the latest news differently? Find us on:

See you tomorrow for a new dose of ☕️ Techpresso!