Anthropic Wins Key U.S. Ruling on AI Training — But It’s Not in the Clear Yet ⚖️

Welcome to another edition of Horizon AI,

Anthropic scored a major victory this week in the battle over using copyrighted materials to train AI models, but they’re not off the hook yet.

Let’s take a closer look!

Read Time: 4.5 min

Here's what's new today in the Horizon AI

• Anthropic Scores Fair Use Win for AI Training, but Faces Trial Over Pirated Books

• Google Enables Imagen 4 in Gemini API and AI Studio

• AI Tutorial: How to use Gemini to identify the name of a song

• AI Tools to check out

• AI Findings/Resources

• The Latest in AI and Tech 💡

AI News

ANTHROPIC

Anthropic Scores Fair Use Win for AI Training, but Faces Trial Over Pirated Books

In a landmark decision, a federal judge ruled that Anthropic’s use of legally purchased physical books to train its AI models qualifies as fair use. However, the company still faces legal trouble for allegedly storing and using pirated books in its training library.

Details:

• Judge William Alsup ruled that digitizing legally purchased books to train AI models qualifies as fair use and is “exceedingly transformative,” noting that the goal is to learn from, not copy, the original texts.

• He compared AI training to educating students and emphasized that the Copyright Act “seeks to advance original works of authorship, not to protect authors against competition.”

• The problem arises from allegations that Anthropic pirated “millions” of books from the internet, which the court found is not protected under fair use. A separate trial will determine potential damages related to this alleged infringement.

• The decision also does not address whether AI-generated outputs infringe copyright, which is a central issue in other ongoing cases.

Overall, the ruling marks a significant win for AI companies and sets a precedent that may influence how judges respond to similar cases going forward.

Source

GOOGLE

Google Enables Imagen 4 in Gemini API and AI Studio

Google has rolled out its latest text-to-image models, Imagen 4 and Imagen 4 Ultra, through the Gemini API (paid preview) and Google AI Studio (limited free testing).

Details:

• Imagen 4 can generate up to four 1024×1024 images per call and supports prompts up to 480 tokens.

• Pricing starts at $0.04 per image, while the Ultra version costs $0.06 and offers tighter prompt alignment and optional 2K exports via Vertex AI.

• Both versions include SynthID watermarks and support near real-time generation.

The Imagen 4 models are among the best AI image generators on the market right now, and their availability in Google AI Studio is a great opportunity for creators to explore new ways of generating high-quality visuals for free.

Source

AI Tutorial

How to use Gemini to identify the name of a song

The Gemini app on Android can now identify songs with help from Google Search.

1. Open the Gemini app.

2. Tap the “Search a song” button at the bottom of the Gemini assistant. A pulsating sphere will appear to show the app is listening. You can also type “What song is this?” to access the feature.

3. You can either sing, hum a tune, or play the song loudly enough in the background—Gemini will compile a list of potential matches, ranked by how closely they match.

Note: This feature appears to be available only on Android for now, and not yet on iOS.

AI Tools to check out

🚀 Rocket: Build web & mobile apps 10x faster without code.

🎬 Descript: AI powered video editor.

👀 Unwink: Transform your data
into actionable insights.

🔍 TwelveLabs: Search and understand your videos - with AI.

📢 Make.ad: Turn text into ads that convert instantly.

AI Findings/Resources

🎓 How ChatGPT and other AI tools are changing the teaching profession

👉 3 common apps you can replace with ChatGPT, and why you definitely should

🌸 How AI infiltrated perfume

🧠 AI vs Human thinking: How LLMs really work

🌐 How to save money for travel using ChatGPT

The latest in AI and Tech

Applebee’s and IHOP have plans for AI

Dine Brands, the parent company of the two chains, plans to use AI in its restaurants and behind the scenes to streamline operations and encourage repeat customers. This includes adding AI-infused tech support for all its franchisees, as well as an AI-powered “personalization engine” to offer customized deals to diners.

ElevenLabs launches a stand-alone voice-generation app

Users no longer need to rely on the web app to use ElevenLabs' tech to generate voice clips from text. It's now available on the mobile app for both Android and iOS.

Is OpenAI entering the self-driving race?

On the “Uncapped With Jack Altman” podcast last week, Sam Altman claimed, “We have some new technology that could just do self-driving for standard cars way better than any current approach has worked.”

OpenAI CEO calls IYO’s trademark lawsuit "silly, disappointing and wrong"

Sam Altman commented on the ongoing trademark lawsuit on X. He also added that IYO CEO Jason Rugolo had asked OpenAI to acquire his company again just days before the lawsuit, despite an earlier rejection.

That’s a wrap!

Thanks for sticking with us to the end! Let’s stay connected on LinkedIn and Twitter.

Not subscribed yet? Sign up here and send it to a colleague or friend!

See you in our next edition!

Gina 👩🏻‍💻