Wyecliff Weekly | April 3th, 2026

Your weekly AI briefing | Edition #17 | Friday, April 3, 2026 OpenAI just raised $122 billion, larger than the GDP of 130 countries, in a single funding round. In the same week, Oracle cut up to 30,000 jobs to pay for AI data centers, Slack rebuilt its product around AI agents, and Anthropic leaked internal secrets twice. The money flowing into AI infrastructure right now is unlike anything the technology industry has seen before, and this week the scale of it has become impossible to ignore. 📰 TOP STORIES + OpenAI Raised $122 Billion and Is Now Worth $852 Billion OpenAI closed the largest private funding round in history on March 31. The company raised $122 billion at a valuation of $852 billion. Amazon led with $50 billion. Nvidia and SoftBank each put in $30 billion. For the first time, OpenAI opened the round to individual investors, raising $3 billion from retail participants. This happened despite months of public backlash, including the #QuitGPT boycott that surpassed 4 million actions and ongoing fallout from the Pentagon deal. OpenAI is generating $2 billion in revenue per month and an IPO is expected as early as Q4 2026. What it all means: OpenAI now has more capital than most countries spend on infrastructure in a decade, and its biggest investors are locking in long term distribution deals alongside their checks. The products built on top of these models are about to get significantly more capable, and the sales teams behind them are about to get significantly more aggressive.  + Anthropic Leaked Twice in Five Days On March 26, a misconfiguration on Anthropic's website left internal documents publicly accessible, including a draft blog post describing "Mythos," a new model the company calls "the most capable we've built to date." The leaked materials described a system with sharply improved reasoning and coding that Anthropic believes poses unprecedented cybersecurity risks. Mythos is currently in limited testing with early access customers. Then on March 31, Anthropic accidentally published the full source code for Claude Code, its popular AI coding tool. A packaging error exposed the entire codebase to the public, and within hours it was copied and analyzed by thousands of developers online. No customer data was involved, but a similar accident happened with an earlier version of the same tool in February 2025. What it all means: Two leaks in five days is either a serious operational problem or a very effective way to generate buzz about a new model without formally announcing it. The Mythos leak put Anthropic's next generation capabilities into every AI newsletter and news cycle on the planet. Anthropic says both were accidents. Maybe, but the timing and the coverage worked out very well for a company that just happens to be competing for enterprise contracts against an $852 billion rival.   + Slack Dropped 30 AI Features in Its Biggest Update Since the Salesforce Acquisition Salesforce CEO Marc Benioff just unveiled more than 30 new AI features for Slack. New additions include tasks, meeting transciption, and summarization. The biggest news: Slackbot can now work outside of Slack itself. It can monitor your desktop activity, including your calendar, deals, and conversations, and use that context to suggest next steps or draft follow ups. It can also connect to outside tools and services without you leaving the app. What it all means: For teams already on Slack, this changes what the tool can do. For teams evaluating AI workflow tools, Slack just became a more serious option.  + Apple Will Let You Choose Your Own AI Inside Siri Last week we covered Apple's deal to hand Siri's default AI over to Google. This week the story got bigger. Bloomberg reported that Apple plans to open Siri to third party AI chatbots starting with iOS 27. A new "Extensions" system in Settings will let iPhone users select which AI services power Siri. If you have the Claude, Gemini, or ChatGPT app installed, you will be able to route Siri queries directly to those services. What it all means: Apple went from picking one AI partner to building a platform where users choose for themselves. Meaning Google now has to to compete for the most valuable distribution channel in consumer technology: the default assistant on a billion iPhones. 📰 MORE STORIES Oracle Slashes Up to 30,000 Jobs to Fund AI Data Centers. On March 31, Oracle sent 6 AM termination emails to employees across the US, India, Canada, and Mexico. Analysts estimate the cuts will hit between 20,000 and 30,000 workers, roughly 18% of Oracle's global workforce. The reason: Oracle needs the cash to fund a massive AI data center expansion. TD Cowen estimates the layoffs could generate $8 billion to $10 billion in incremental free cash flow. Oracle posted $6 billion in quarterly income last quarter. A Federal Judge Blocked the Pentagon From Banning Anthropic. U.S. District Judge Rita Lin issued a preliminary injunction on March 26, ruling the government's attempt to label Anthropic a "supply chain risk" appeared to be "designed to punish" the company rather than protect national security. The dispute stems from Anthropic's refusal to grant the Pentagon unrestricted access to its AI models for all purposes, including autonomous weapons. The case continues.​ A Popular AI Developer Tool Was Caught Stealing Passwords. LiteLLM, an open source tool used by developers to connect to AI models and downloaded up to 3.4 million times per day, was infected with malware this week. The malware harvested login credentials from anyone who installed the compromised version. 💡 WHY IT MATTERS $122 billion raised by one company. 30,000 jobs eliminated by another. 30 new AI features shipped inside a tool used by millions of workers every day. The largest companies in the world put real money, real headcount, and real product roadmaps behind AI in ways they cannot easily walk back. For You: Apple opening Siri to Claude, Gemini, and ChatGPT is the most consumer relevant story this week. Starting later this year, you will be able to pick which AI powers your phone's assistant. It means the AI you prefer using on your laptop can follow you to your phone, your tablet, and your watch. If you have been frustrated by Siri, this is the update you have been waiting for. For Your Business: The Slack update is the one most likely to affect your day to day. If your team uses Slack, you now have an AI assistant that can summarize meetings, draft follow ups, and connect to other business tools without leaving the app. Separately, OpenAI's funding round means its enterprise tools are about to get more aggressive on pricing, features, and sales outreach. If your company has not been contacted by an AI vendor yet, that is about to change.  🛠️ ONE THING TO TRY THIS WEEK How to Upload and Work With Your Own DocumentsLast week we covered how to set up recurring AI tasks. This week builds on that. You have told AI who you are and what to do on a schedule. Now you give it something real to work with: your own files. Every major AI platform lets you upload documents and ask questions about them. Upload a contract and ask for a summary of key terms. Upload a spreadsheet and ask for trends. The AI reads the document and responds based on what is actually in the file, not general knowledge. Here is how to do it on each platform. Claude (via claude.ai or Cowork) Open claude.ai and start a new conversation. Click the paperclip icon at the bottom of the chat window, or drag and drop a file directly into the conversation. Claude supports PDFs (up to 100 pages), Word docs, Excel files, PowerPoints, CSVs, and images. Type your question after attaching the file. Claude reads the full document, including charts and tables, and responds based on the contents. Bonus: Add files to a Project (left sidebar) and every conversation inside that project will have access to those files automatically. ChatGPT Open chatgpt.com and start a new conversation. Click the paperclip icon or the plus (+) button and select "Add from computer." ChatGPT supports PDFs, Word docs, Excel files, PowerPoints, CSVs, and images. File size limit is 512 MB. Ask your question after attaching the file. Free users can upload 3 files per day. Paid users can upload up to 80 files every 3 hours. Bonus: Use Projects (left sidebar) to store files that persist across conversations without re-uploading. Microsoft Copilot Open copilot.microsoft.com and start a new conversation. Click the plus (+) icon and select "Add images or files." Copilot supports Word docs, PowerPoints, Excel files, and PDFs. Upload and ask your question. For Microsoft 365 users, you can also reference files already in OneDrive or SharePoint by typing "/" and selecting a file. 💬 Try This Prompt Upload a document you are working with this week and try this: "Read this document and give me: 1) A three sentence summary, 2) The three most important details I should not miss, and 3) Two questions I should be asking based on what is in here." The Wyecliff Weekly is brought to you by Wyecliff. Wyecliff helps businesses identify where AI fits into their operations and build it in a way that delivers real results. From Microsoft Copilot deployments to custom automation and workflow integration, we work with companies across construction, distribution, insurance, and beyond. If your team is ready to move from talking about AI to actually using it, we are happy to have that conversation. 🌐 wyecliff.ai