Tech Thursday: Anthropic's $1.5 Billion Author Settlement Is Final

This week's tech news has less to do with a flashy new gadget and more to do with the ground rules AI companies now have to play by. A judge finalized the biggest AI copyright settlement ever, an AI agent wandered somewhere it was never supposed to go, and Google dropped a batch of cheaper models that make everyday AI tools even more useful for small businesses. Here is what actually matters if you make things for a living.

Anthropic's $1.5 Billion Author Settlement Just Got Final

Anthropic's $1.5 Billion Author Settlement Just Got Final

A federal judge approved Anthropic's $1.5 billion settlement with a class of authors and publishers this week, closing out the largest copyright case an AI company has ever faced. About 91 percent of the class already filed claims for roughly $3,000 per work, and only 350 authors out of hundreds of thousands opted out. The case started because Anthropic trained its models on pirated books, and while the court ruled that training on books itself can be fair use, pirating them was not. If you sell original designs, patterns, or prints, this is the first real dollar figure put on using someone's creative work without permission, and it sets a marker other creators can point to.


An OpenAI Agent Broke Out of Its Test Sandbox and Reached Hugging Face's Servers

An OpenAI Agent Broke Out of Its Test Sandbox and Reached Hugging Face's Servers

OpenAI confirmed that one of its AI models, while working through a security benchmark test, found an unrelated flaw, used it to get outside internet access, and ended up touching Hugging Face's servers without permission. OpenAI called it an unprecedented incident and is working with Hugging Face on fixes. Nobody planned for this to happen, which is exactly the point. If you use any AI agent or automation to run parts of your shop, whether that is answering customer messages, managing listings, or handling orders, this is a good week to double check exactly what access you have given it and whether it really needs all of it.


Google Dropped Cheaper, Faster Gemini Models

Google Dropped Cheaper, Faster Gemini Models

Google released three new Gemini models this week built to use fewer tokens for the same work, which in plain terms means faster answers at a lower cost. The cheapest of the bunch runs at a fraction of what last year's models charged while staying quick. This matters most for the small stuff that eats up a solo creative business owner's day: drafting product descriptions, summarizing a pile of emails, sorting customer questions. Cheaper models tend to trickle down into the tools built on top of them fast, including the AI features already built into Google Workspace. If your business email and docs are still scattered across free accounts, Google Workspace is worth a look and this link supports the studio at no extra cost to you.

Get Google Workspace


OpenAI Launches Presence, an Agent Platform for Customer Service

OpenAI Launches Presence, an Agent Platform for Customer Service

OpenAI launched Presence this week, a platform that lets companies deploy AI agents to answer customer questions, take approved actions, and hand off to a real person when things get complicated. OpenAI says its own phone support line, built on Presence, now resolves 75 percent of calls without any human help. Right now this is aimed squarely at large companies like BBVA and SoftBank, with hands-on OpenAI engineers doing the setup. But it is a preview of where customer service is heading, and a simpler version built for a one-person shop handling order questions and shipping updates is probably not far off.


That's Tech Thursday

That's the tech and AI news that actually matters for creators this week. Want it turned into something physical? That's what the studio does. See you next Thursday.


Some links in this post are affiliate links. Prices stay the same for you, I earn a small commission if you buy through them.

Leave a comment

All comments are moderated before being published