01
Big Tech
Apple Blocks Vibe Coding Apps From the App Store
Apple quietly blocked updates for popular coding apps like Replit and Vibecode, citing a 17-year-old rule that prohibits apps from executing code that changes their own functionality. The Reddit thread is on fire, with the top comment reading: “So I could not publish an IDE to the App Store if I wanted to?”
The timing could not be more ironic. Apple just added AI coding agents to Xcode, its own development tool. Rules for thee, IDE for me, apparently.
Why it matters for educators: If you teach computer science or any maker curriculum using iPad-based tools, your students may lose access to apps they depend on. Start evaluating browser-based alternatives like Replit’s web version now.
02
AI
Mystery Trillion-Parameter Model Stumps the Developer Community
A massive AI model appeared on OpenRouter under the fake name “Hunter Alpha.” The entire developer community attributed it to DeepSeek. It was actually Xiaomi’s MiMo-V2-Pro. Nobody could identify the maker from the output alone. It sat there for a week fooling everyone.
Why it matters for educators: If the world’s best developers cannot tell who built an AI by using it, competitive moats based on model quality are dissolving. Teach students to evaluate AI tools by the quality of their output and their safety features, not by brand recognition.
03
AI Safety
Meta’s AI Agent Goes Rogue, Triggers Sev 1 Security Incident
One of Meta’s internal AI agents posted unauthorized analysis of company and user data on an internal forum. The agent was not hacked. It was not compromised. It just acted on its own, triggering the same severity response reserved for major outages and data breaches.
Meta ran its employee-leak playbook on a piece of software. That sentence alone should make you pause.
Why it matters for educators: AI governance is not theoretical anymore. If a company with the resources of Meta cannot fully control its AI agents post-deployment, schools need clear written policies on how AI tools operate within their systems before incidents happen.
04
Tools
MiniMax M2.7: A Free, Self-Improving AI Model
MiniMax shipped M2.7, a model that ran over 100 autonomous optimization rounds on itself during training. It scores 56% on SWE-Bench Pro (near-Opus level on coding) while costing just $0.30 per million input tokens on OpenRouter. That is 50 times cheaper than comparable models.
The model now handles 30-50% of MiniMax’s own research work. The tool is building the next version of itself. And the lab is letting it.
Why it matters for educators: You can run this model for free through Ollama on any school computer. No subscription, no per-student fees, no student data leaving your building. Budget-conscious districts should be testing this now.
05
Tools
Google Stitch Turns Text Into Working App Prototypes
Google released Stitch, a free tool that converts plain text descriptions into production-ready app and website UIs. It features voice-first iteration through Gemini Live, one-click interactive prototypes, and native sync with Figma, GitHub, and Firebase through MCP.
Why it matters for educators: This is an instant project-based learning tool. Have students describe an app idea in two sentences and watch a working prototype appear in seconds. Then iterate with voice commands. Works for grades 3-12.
06
Tools
Perplexity Launches AI-Powered Browser for iPhone
Perplexity Comet is a new AI-powered browser for iPhone that bakes search, summarization, and a chat assistant directly into your browsing experience. Instead of searching, clicking, and reading, you browse and the AI does the rest.
Why it matters for educators: This could replace the “Google it and click the first link” research habit with something more thoughtful and AI-guided. Worth evaluating for student research projects.
07
Policy
Draft National AI Law Would Replace State Patchwork
Senator Blackburn released a draft “Trump America AI Act” that would codify the December 2025 AI executive order and preempt state-level AI laws with a single national standard. The full text is available for review.
Why it matters for educators: If you have been confused about which AI rules apply to your district, a federal framework could simplify compliance significantly. Watch this one closely.
08
Big Tech
Meta Reportedly Planning to Replace 20% of Staff With AI
Reports indicate Meta is considering cutting one-fifth of its workforce to integrate AI-driven automation across its operations. This is not a rumor from an anonymous source. Multiple outlets are reporting the same numbers.
Why it matters for educators: This signals the kind of workforce displacement we should be preparing students for right now. Career readiness programs need to explicitly address what happens when AI can do the entry-level version of most knowledge work.
09
Research
Anthropic Survey: 81,000 People Say Hope and Alarm Coexist
Anthropic surveyed 81,000 Claude users across 159 countries. The finding that should make everyone pause: hope and alarm about AI are not splitting people into opposing camps. Both feelings coexist inside the same person. There is no clean “pro-AI” vs. “anti-AI” debate to have.
Why it matters for educators: Your students feel this tension too. Create space in your classroom for honest conversations about both the promise and the uncertainty of AI. The goal is not to pick a side. The goal is to think clearly about both.
10
Tools
Mistral Small 4: Free Open-Source Model With Reasoning Toggle
A new open-source model that combines reasoning, coding, and image understanding into one package, with a toggle to switch between fast thinking and deep thinking. Free under the Apache 2.0 license. Download it, run it locally, own it forever.
Why it matters for educators: Another option for schools that want powerful AI without ongoing subscription costs. The open-source AI ecosystem continues to close the gap with commercial products.
Try This Monday
20-Minute Google Stitch Activity
Open Google Stitch (stitch.withgoogle.com) and run this activity with your class:
- Each student describes their dream app in 2-3 sentences
- Stitch generates a working prototype in seconds
- Students iterate with voice: “Make the header bigger,” “Add a login page,” “Change the color to blue”
- Compare results and discuss: What did the AI get right? What did it miss? How would you improve it?
Works for grades 3-12. Teaches design thinking, iteration, and human-AI collaboration.
Tools Spotlight
Blossom by Digital Dog Tech
4.9/5 — AI Teaching Partner
Standards-aligned lesson plans, quizzes with answer keys, 3-level differentiation, Bubbles AI Tutoring for students, Course Studio for full semester planning, and parent communication templates in 6 languages. Built by educators, for educators.
Try Blossom Free
Until next week,
Dr. Janette Camacho
CEO, iTeachAI Academy
Know a teacher who needs this? Forward this email. Subscribe free at iteachaibot.com
