The AI Dispatch — April 9, 2026

Platforms & Products

Google Folds NotebookLM Into the Gemini App, Turning Chat Into a Research Operating System

Subscribers to Gemini Ultra, Pro, and Plus can now create persistent “Notebooks” directly inside the Gemini app — syncing bidirectionally with full NotebookLM workspaces for the first time.

Sources: Google Blog • 9to5Google • Dataconomy

Google has been running two parallel knowledge products for over a year — the conversational Gemini app and the research-heavy NotebookLM — and the seam between them was starting to show. On Wednesday the company eliminated it. Gemini subscribers can now tap a “Notebooks” tab inside the main app, creating persistent workspaces that organize chats, uploaded PDFs, Google Drive files, URLs, and other source material into structured research repositories. Those repositories sync bidirectionally with standalone NotebookLM, meaning anything added in one surface appears instantly in the other.

The integration also brings NotebookLM’s most distinctive output formats — Video Overviews, auto-generated Infographics, and study guides — into the Gemini conversation flow. A user can ask Gemini a question, pin the response to a Notebook, pull in three supporting PDFs from Drive, and generate a video summary without ever leaving the app. For researchers and students who had been tab-switching between the two products, the merge removes a friction point that had long felt artificial.

The move also signals Google’s broader product strategy: rather than shipping standalone AI tools, it is collapsing them into a single surface that doubles as both assistant and workspace. Whether that convergence delights power users or overwhelms casual ones is the open question the next quarter of usage data will answer.

Pricing & Developer Tools

OpenAI Launches $100/Month ChatGPT Pro Tier, Anchored by Codex

A new mid-range plan fills the gap between the $20 Plus and $200 tiers, giving power users and developers substantially more coding capacity per dollar.

Sources: TechCrunch

OpenAI introduced a $100-per-month ChatGPT Pro plan on Wednesday, explicitly targeting the developer and power-user segment that had been forced to choose between the modest limits of the $20 Plus subscription and the steep $200 ceiling of the top tier. The new plan centers on Codex, OpenAI’s code-generation engine, delivering what the company describes as meaningfully higher throughput for agentic coding tasks — longer sessions, more parallel tool calls, and priority access during peak demand.

The pricing move is as much about market positioning as revenue. At $100 a month, OpenAI is attempting to set the reference price for professional-grade AI coding assistance — a figure low enough to undercut hiring a junior contractor but high enough to signal that serious tooling is not free. With Anthropic’s Claude Code, Google’s Gemini Code Assist, and GitHub Copilot all competing for the same developer wallet, the mid-tier plan is a bet that volume at a lower price point will capture more seats than the premium tier alone ever could.

Industry & Business

Mobile & Consumer AI

Meta AI App Surges to No. 5 on App Store After Muse Spark Launch

The Meta AI standalone app climbed to fifth place on the Apple App Store charts within twenty-four hours of Meta’s Muse Spark debut. The timing is not coincidental: Muse Spark, a proprietary image and short-video generation model unveiled Tuesday, drove a wave of downloads from users eager to try the creative features before competitors could replicate them. The chart surge validates Meta’s decision to keep its most capable consumer-facing model proprietary rather than releasing it under the Llama open-weights umbrella — a strategic fork that has drawn both praise from investors and criticism from the open-source community.

Sources: TechCrunch

Revenue & Growth

Perplexity ARR Hits $450M After 50% Single-Month Surge

Perplexity AI’s annualized recurring revenue reached $450 million in March 2026, a 50 percent increase in a single month. The acceleration coincides with the company’s pivot from pure search to a broader AI agent and autonomous-computer strategy, which has pushed monthly active users past the 100-million mark. The revenue trajectory — from roughly $100 million ARR in late 2025 to $450 million barely six months later — makes Perplexity one of the fastest-scaling enterprise AI companies on record, though investors note that sustaining this growth rate will require the agent products to retain users beyond the initial novelty window.

Sources: The Rundown AI

Copyright & Litigation

RIAA and Music Industry Bodies File Amicus Brief in Anthropic Copyright Suit

The RIAA, NMPA, A2IM, SONA, the Black Music Action Coalition, and SoundExchange jointly filed an amicus brief supporting music publishers’ motion for partial summary judgment in their $3.1 billion copyright suit against Anthropic. The brief alleges that Anthropic acquired over 20,000 copyrighted works — including “Wild Horses,” “Sweet Caroline,” and “Eye of the Tiger” — via pirate library sites to train Claude. The filing represents the broadest coalition of music-industry organizations to weigh in on an AI training-data case to date, and its framing positions the suit as a test case for whether mass ingestion of copyrighted creative works constitutes fair use under U.S. law.

Sources: Music Ally

Legal & National Security

DC Circuit Fast-Tracks Anthropic Pentagon Appeal

After denying Anthropic’s emergency request to pause a Pentagon supply-chain risk designation, the D.C. Circuit Court of Appeals agreed to fast-track the full appeal. The expedited schedule signals that the court views the underlying legal questions — whether national-security agencies can unilaterally designate AI companies as supply-chain risks, and what due-process protections apply — as significant enough to warrant resolution before the designation’s practical effects become entrenched. The case is being closely watched across the defense-technology sector, where several other AI firms face similar classification reviews.

Sources: National Law Journal

Quick Dispatches

134 AI-in-Education Bills Introduced Across 31 States

A MultiState analysis documented 134 active AI-in-education bills spanning 31 state legislatures, with the heaviest concentration on student data privacy, classroom use boundaries, and curriculum requirements. California’s AB 1159 would prohibit using student data to train AI models; Georgia and Mississippi now require AI literacy instruction for high-school graduation. The legislative wave marks the fastest state-level policy response to any technology category since social-media regulation in 2023.

MultiState

Revolut Launches AIR Financial Assistant

Revolut debuted “AIR,” an AI assistant that handles daily financial tasks through natural-language chat — account inquiries, transaction categorization, spending insights, and bill reminders. The launch puts Revolut alongside Chase and Bank of America in the growing cohort of consumer banks deploying conversational AI as a first-class interface, rather than relegating it to FAQ chatbots buried in help menus.

Idaho Signs Chatbot Safety Act into Law

Idaho’s Conversational AI Safety Act (SB 1297) was signed into law, requiring chatbot operators to provide parental management tools for users under 13, including content filtering, session time limits, and data deletion on request. Oregon signed a similar measure the same week, and Nebraska’s companion bill is advancing through committee — establishing a clear regional pattern of child-safety requirements for conversational AI products.

JDSupra