Welcome back.
Today, AI is colliding with law, hardware and platform lock-in at the same time. This edition is an investigation into who controls the stack, how technology is entering its AI-native phase, and which movers and shakers are adding hot new technology capabilities to your daily AI use.
In today’s Generative AI Newsletter:
The U.S. Supreme Court refuses to hear the biggest AI copyright case yet
Apple launches $599 devices built for on-device Apple Intelligence
Anthropic makes it easier to move your AI “memory” into Claude
Alibaba’s compact Qwen3.5 models punch far above their weight
Latest Developments
Supreme Court ducks the AI authorship question
The U.S. Supreme Court declined to hear the leading case on whether AI-generated art can be copyrighted. That leaves lower court rulings in place that say copyright requires a human author.
The case in summary:
The case stems from computer scientist Stephen Thaler, who tried to register copyright for artwork generated by his AI system, DABUS.
The Copyright Office rejected the application.
Courts backed the “humans only” standard, and the Supreme Court has now chosen not to revisit it.
The practical reality is messy. Pure machine-generated work remains uncopyrightable. AI-assisted work still depends on how human contribution is framed and evidenced. This will come back, likely with much bigger players and much higher stakes.
Special highlight from our network
Most AI talks.
Anam’s Interactive Avatars connect.
With photorealistic rendering and expressive emotional cues, users choose avatars 70% over voice or text alternatives. That preference drives higher usage, stronger engagement, and better outcomes across sales, onboarding, and customer support.
Designed for real products:
• 99.95% uptime SLA
• Sub-180ms response time
• High concurrency support
• Real-time interruptibility
• Low-code integration
Built for flexibility:
Generate a persona from a single image. Plug in your own LLM. Clone and customize voice. Define behavior by scenario.
When AI represents your brand, realism matters.
Attio is the AI CRM for modern teams.
Connect your email and calendar and Attio instantly builds your CRM. Every contact, every company, every conversation — organized in one place. Then ask it anything. No more digging, no more data entry. Just answers.
What If Your AI Couldn’t Lie to You?
AI gives you better results when it’s connected to real, up-to-date data. But access alone isn’t enough.
MongoDB’s document model brings structured and unstructured data together, making it easier to power accurate, context-aware RAG applications with vector search.
The result: smarter apps you can trust.
Learn how to build it step by step.
Apple pushes on-device AI into the mainstream price tier
Apple just announced the iPhone 17e and a refreshed iPad Air, both starting at $599 and positioned to run Apple Intelligence on-device.
The iPhone 17e includes the A19 chip, 256GB base storage, and AI features like Call Screening, Hold Assist, and Live Translation.
The iPad Air moves to the M4 chip with 12GB unified memory, giving more headroom for AI workloads.
Pre-orders begin March 4. Shipping starts March 11.
The message is clear. AI is becoming a default expectation in consumer hardware, not a premium add-on. The question is: will Apple pay for dallying and keepign consumers waiting for on-device AI?
Anthropic wants to make switching AI assistants frictionless
Anthropic just released a tool that lets users port saved preferences and context from other AI providers into Claude using a copy-paste flow.
It pulls items like:
saved instructions
personal preferences
project context
behavioural settings
…from systems like ChatGPT, Gemini, and Copilot, then imports them into Claude’s memory.
Anthropic also expanded memory access to free users, and added stronger persistent memory for Claude Code.
This is the platform war moving up a layer. Model quality matters. But user context is becoming a huge retention lever.
Alibaba’s small models raise the bar for “local AI”
Alibaba just released Qwen3.5 Small, a set of open-source models designed to run on phones and laptops, ranging from 0.8B to 9B parameters.
Alibaba says the 9B model outperforms an OpenAI open model at 120B parameters on certain reasoning and multilingual benchmarks. The series supports text, images, and video, and is licensed for commercial use.
The big change here is not about replacing frontier models. It’s making capable AI cheap, portable and deployable without a cloud bill. That is how adoption spreads.
