Plus: First countries block Musk's AI over deepfakes, Character.AI settles teen suicide suits, and Anthropic angers developers
Sponsored by
Create full-stack, production-ready web and mobile apps from natural language prompts with Emergent — the first agentic vibe-coding platform that handles everything from planning to deployment.
Learn MoreThis week was a wake-up call for the entire AI industry. Malaysia and Indonesia became the first countries to outright ban Grok after a tsunami of nonconsensual deepfakes targeting women and children. The UK is threatening to follow with potential fines or a full ban. Character.AI quietly settled lawsuits from families whose teens died by suicide after chatbot interactions. Meanwhile, Anthropic sparked developer outrage by blocking third-party apps from using Claude subscriptions, then dropped Cowork, a new agent that basically built itself in a week. Oh, and AI-generated images of Maduro's capture may have outnumbered the real footage. Wild times.
On this day in 2025…
President Biden issued Executive Order 14141 on "Advancing United States Leadership in AI Infrastructure." The order directed federal agencies to lease lands for AI data centers and streamline environmental permits, setting the stage for America's AI infrastructure boom. One of Biden's final AI policy moves before leaving office.
🚫 Countries Block Grok Over Deepfake Scandal
Malaysia and Indonesia became the first nations to ban Musk's Grok AI after a flood of nonconsensual sexualized deepfakes, including images of women and minors. The UK's Ofcom launched a formal investigation, with PM Starmer calling the content "disgusting" and "unlawful." X could face fines up to 10% of global revenue or an outright ban.
🔒 Anthropic Blocks Third-Party Apps from Claude Subscriptions
Anthropic cracked down on developers using Claude Max subscriptions through tools like OpenCode and Cursor, triggering widespread backlash. The $200/month unlimited access loophole is now closed, forcing heavy users to the metered API. Rails creator DHH called it "very customer hostile" as developers canceled subscriptions en masse.
🤖 Anthropic Unveils Claude Cowork
Anthropic launched Cowork, a file-managing AI agent that brings Claude Code's power to non-developers. Give it access to a folder, and it reads, edits, and creates files on your behalf. The kicker? The team built the entire tool in about a week and a half using Claude Code itself. Currently a research preview for Max subscribers ($100-$200/month).
⚖️ Character.AI Settles Teen Suicide Lawsuits
Google and Character.AI reached settlements in five lawsuits alleging chatbots contributed to teen suicides and mental health crises. The landmark cases include Megan Garcia's suit over her 14-year-old son who died after conversations with a "Game of Thrones" chatbot that encouraged self-harm.
🎭 AI Deepfakes Flood Venezuela Coverage
Fake AI-generated images of Maduro's capture racked up over 14 million views within 48 hours on X. Even Musk shared what appeared to be AI-generated celebration videos. Experts say there may have been more fake content than real footage, marking a new era in real-time misinformation.
🏆 CES 2026 "Worst in Show" Roasts AI Overload
Consumer advocates handed Samsung's AI fridge the top anti-award for voice-controlled doors that fail in noisy kitchens, built-in ads, and always-on cameras. Amazon Ring's expanding surveillance features and a disposable electronic lollipop also made the list of tech that "undermines privacy, security, and repairability."
⚡ LocalAI is the blazing-fast, self-hosted open-source alternative to proprietary APIs like OpenAI and Claude—running multimodal inference (text, audio, video, images, voice cloning) entirely locally on consumer hardware with no GPU required, featuring P2P distributed workloads and drop-in compatibility for seamless migrations.
🖥️ Beowulf AI Cluster turns mismatched everyday hardware (old PCs, Raspberry Pis, laptops) into powerful distributed AI clusters using Ansible, enabling benchmarked inference with tools like llama.cpp and Exo—perfect for cost-free scaling of local LLMs without cloud dependency.
🔧 AlpaSim is NVIDIA's newly released open-source end-to-end simulation framework for high-fidelity autonomous vehicle development, offering realistic sensor modeling, traffic scenarios, and scalable testing to accelerate safe AI-driven mobility research.
🚀 Alpamayo delivers NVIDIA's groundbreaking open-source vision-language-action reasoning models and datasets tailored for autonomous vehicles, bridging interpretable planning with precise control—now available for community fine-tuning and advancement in physical AI.
Coming in Hot
Privacy-first text analyzer with offline editing.
AI writing platform that turns ideas into content.
AI product image generator for e-commerce.
AI browser agents that automate web tasks 24/7.
AI coding agent that handles dev tasks end-to-end.
AI agent that automates research, workflows, and email
AI anime and video creation studio in one platform.
AI-assisted project planner for developers.
AI writing assistant that fixes grammar instantly.
Voice-driven AI workspace for smarter productivity.
Submit your AI tool to the #1 AI tools marketplace.
The simplest free setup that actually works
You do not need cloud IDEs, expensive servers, or complicated setups to code from your phone.
All you need is your existing Mac, a secure VPN, and a proper SSH client.
This setup takes about 10 minutes and costs nothing.
• Secure SSH access to your Mac from anywhere
• A smooth mobile typing experience with shortcut keys
• Your real local dev environment on your phone
• No public ports, no router config, no fear
1. Tailscale for VPN and security
2. Termius for SSH access with a great mobile keyboard bar
3. A Mac that stays powered on with the lid open
That is it.
Tailscale creates a private VPN between your devices using WireGuard.
1. Go to Tailscale and download the macOS app
2. Sign in using Google, GitHub, or Apple
3. Turn Tailscale on
Once enabled, your Mac now has a private IP like:
100.x.x.x
This IP is only accessible by your own devices. No public internet exposure.
1. Install Tailscale from the App Store or Play Store
2. Sign in with the same account
3. Turn it on
Your phone and Mac are now on the same private network, even if you are on mobile data.
On your Mac:
1. Open System Settings
2. Go to General → Sharing
3. Turn on Remote Login
4. Make sure your user is allowed
You can test locally with:
ssh your_username@localhost
If that works, you are good.
Termius is an SSH client designed for mobile.
The extra keyboard row alone makes it worth it.
1. Install Termius
2. Create a new Host
3. Fill in:
• Host: your Mac’s Tailscale IP
• Username: your macOS username
• Authentication: Password or SSH key
Save it.
For this to work smoothly:
• Plug your Mac into power
• Keep the lid open
• Disable sleep or set it to never sleep
You can do this in System Settings → Lock Screen.
Optional but recommended:
Use Amphetamine to keep the Mac awake.
Now open Termius and connect.
You are inside your real Mac terminal.
From here you can:
• Run git pull
• Start npm run dev
• Edit files with vim, nano, or tmux
• Control Docker, servers, or scripts
• Push commits from your phone
All from the beach, train, or couch.
Why this setup is so good
• No exposed SSH ports
• No port forwarding
• No cloud costs
• Uses your real dev environment
• Works on WiFi, 5G, or public networks
It feels like cheating, but it is not.
• Use SSH keys instead of passwords
• Enable Touch ID for sudo on macOS
• Use tmux so sessions survive disconnects
• Pair with GitHub CLI for full workflow
This is the simplest free way to code from your phone without compromises.
Once you try it, you will never look at cloud IDEs the same way again.
Now you can SSH anytime and start vibing from your phone.
🔍 The Great AI Coding Credit Debate
Developers in 2026 are asking a new question: "Which tool won't torch my credits?" Pricing models are now debated as intensely as capabilities, with 65% of developers using AI coding tools weekly according to Stack Overflow's 2025 survey.
🛡️ AI Safety Gets a Key Update
The International AI Safety Report released a critical update in January 2026, noting that AI capabilities are moving so fast that yearly reports can't keep pace—changes now happen on a timescale of weeks, not months.
⌨️ TypeScript Overtakes Python on GitHub
TypeScript became the most-used language on GitHub as of August 2025, overtaking both Python and JavaScript with over 1 million new contributors in 2025. Evidence suggests AI-assisted development boosted TypeScript's rise, changing not just how fast code is written, but which languages developers choose.
🎯 GitHub Copilot Goes Full Agent
GitHub rolled out Agent Mode with Model Context Protocol support to all VS Code users, transforming Copilot from code completion to a fully agentic development partner. Unlike traditional chat, Agent Mode independently translates ideas into code, identifying subtasks and executing across multiple files.
🧠 Anthropic's "Do More With Less" Strategy
Daniela Amodei told CNBC that Anthropic has always had a fraction of competitors' compute and capital, yet consistently delivered the most powerful models. She argues the next phase won't be won by the biggest pre-training runs, but by who delivers the most capability per dollar of compute.
📊 a16z Raises $15 Billion in New Funds
Andreessen Horowitz announced $15 billion in new funds on January 9, 2026—its largest fundraising haul to date. The firm promises to back startups in AI, crypto, and American Dynamism as venture dollars continue flowing heavily into AI companies.
