Happy Friday, fellow AI watchers! 🤖

Another week, another dozen "this changes everything" announcements. Let's sort the signal from the noise.

🔥 Top Stories This Week

1. Claude Gets App Connectors (and Actually Means It) Anthropic just rolled out app connectors to all Claude users — including mobile beta. Not "coming soon," not "limited access for enterprise tier 4 subscribers." All of you. Claude can now hook into your apps and do actual things. Wild concept: an AI assistant that assists.

2. Google's 75% Moment Sundar Pichai casually dropped that 75% of all new Google code is now AI-generated — up from 50% last fall. For context, Anthropic's Claude Code already writes 70–90% of Anthropic's own code. We are basically watching software eat itself and write its own replacement simultaneously.

3. ChatGPT Images 2.0 is Unhinged (In a Good Way) OpenAI dropped Images 2.0 this week with multilingual text, full infographics, slide decks, maps, and manga. Seemingly flawlessly. Designers are either very excited or furiously updating their LinkedIn bios.

4. Google Deep Research Gets an Enterprise Upgrade Built on Gemini 3.1 Pro, the new Deep Research and Deep Research Max agents can now search both the web and your private data simultaneously. Finance and life sciences folks: your junior analyst is now an AI that doesn't need coffee breaks or health insurance.

5. Nothing Launches Essential Voice The Phone maker (yes, that Nothing) launched Essential Voice — a dictation tool covering 100+ languages that tidies up your speech in real time. Small launch, but it's signaling something: voice-first interfaces are creeping back, and this time they might actually work.

📈 Surprising Trend: The Free Ride Is Officially Over

Ads. Rate limits. Feature restrictions. Price hikes. As one headline put it this week: "The AI free ride is over." Every major platform is tightening the belt on freemium tiers while simultaneously promising the premium stuff is totally worth it. The great AI land-grab is transitioning into the great AI monetization scramble. Buckle up.

😂 Hot Take of the Week

A Big Law firm — Sullivan & Cromwell, fancy enough to represent SpaceX's merger — just had to apologize to a federal judge for filing briefs full of AI-hallucinated fake case citations. The list of errors ran three pages long. Three. Pages.

Somewhere, a very tired associate is updating their résumé. Somewhere else, an AI is confidently citing Fictional Corp v. Made-Up LLC, 2019.

The lesson: AI is a power tool, not a paralegal. Please supervise accordingly. ⚖️

That's a wrap for the week. Stay curious, stay skeptical, and maybe double-check your AI's citations before filing them in federal court.

Until next Friday, The AI Tool HQ Team 🛠️

You're receiving this because you signed up for AI Tool HQ updates. Unsubscribe anytime — no hard feelings.