AI training gets legal clarity

Good morning, AI enthusiasts. AI’s fair use debate just got its first major answer — and it’s each a win and a warning for AI corporations.

With a judge green-lighting Anthropic’s use of legally purchased data but slamming them for using over 7M pirated books, the split decision is likely to be a robust (but costly) precedent for AI’s legal woes.

Reminder: Our next live workshop is today at 3 PM EST — join and learn easy methods to construct production-grade AI agents that mix context, reasoning, and actions with Gumloop. RSVP here.

In today’s AI rundown:

• Anthropic scores win over AI ‘fair use’ claim

• OpenAI’s Workspace, Office competitor

• Find out how to automate content strategy with scheduled tasks

• LinkedIn co-founder bets on AI ultrasound helmet

• 4 latest AI tools & 4 job opportunities

LATEST DEVELOPMENTS

ANTHROPIC

📚 Anthropic scores win over AI ‘fair use’ claim

The Rundown: A federal judge just ruled that Anthropic’s AI training on legally purchased books qualifies as fair use, but rejected any defense for 7M pirated copies in its digital library in the corporate’s class motion lawsuit filed by a bunch of authors.

The small print:

• The judge called AI training “spectacularly” transformative, comparing Claude to aspiring writers learning from established authors reasonably than copying them.

• The authors did not display that Claude could generate outputs resembling their original works, weakening core claims about competitive harm.

• The filings revealed that Anthropic legally spent “many thousands and thousands” to buy print books, scanning them into digital files to be used in AI training.

• Nevertheless, Anthropic also downloaded thousands and thousands of books from pirate sites, storing them permanently, which the court said violated authors’ rights.

• The corporate will face trial in December for willful infringement of the pirated works, with potential damages potentially reaching $150,000 per book.

Why it matters: AI labs get a green light to coach on legally obtained data, which could provide considered one of the primary precedents for countless cases currently levied against tech corporations. Nevertheless, given the dearth of clarity around what number of copyrighted materials get into training data, this remains to be only one battle in what looks like a really long legal war.

TOGETHER WITH H COMPANY

🤖 The AI agent that completes tasks for you

The Rundown: Backed by a historic $220M seed round, H Company just released their flagship Runner H AI agent that permits you to automate entire tasks in a number of seconds — now available in beta.

Some examples of tasks you may delegate to Runner H:

• Reading your essential emails and drafting (and even sending!) replies

• Making a Google Sheets with trending ad ideas and sharing it together with your team on Slack

• Finding job opportunities and applying for you

Offload some work to Runner H today — try it now totally free.

OPENAI

📊 OpenAI’s Workspace, Office competitor

The Rundown: OpenAI is reportedly constructing productivity tools for ChatGPT that mirror Google Workspace and Microsoft Office — with features like real-time document collaboration and multi-user chat.

The small print:

• OpenAI CPO Kevin Weil reportedly first showcased collaboration designs last yr, though development stalled until the Canvas interface launch in October.

• The Information reports that OpenAI has built but has yet to release multiuser chat, allowing teams to speak about shared work directly in ChatGPT.

• OpenAI also recently rolled out a record mode for transcriptions, file uploads to Projects, and connectors to drag data from Teams, Drive, and DropBox.

• Business subscriptions generated $600M in 2024, with OpenAI projecting $15B by 2030, with increased revenue coming from enterprise subscriptions.

Why it matters: Sam Altman warned last yr that OpenAI would “steamroll” most AI startups… But he may also have his biggest partner within the crosshairs. ChatGPT’s productivity push is about to step right on Microsoft’s legacy software — and given the icy current relationship, the renegotiation may get much more contentious.

AI TRAINING

🤖 Find out how to automate your content strategy with scheduled tasks

The Rundown: On this tutorial, you’ll learn easy methods to use Grok’s latest Tasks feature to schedule automated content research and trend evaluation — running silently within the background and notifying you once it’s done.

Step-by-step:

1. Visit Grok and navigate to the “Tasks” section

2. Click “Add latest” and create your first task: “Weekly Content Trends Evaluation” scheduled for Sundays

3. Use this prompt: “Analyze trending topics in [your niche] from the past week. Discover 5 content opportunities with content format recommendations”

4. Create additional tasks similar to viral content evaluation, or monthly market research (as much as 10 total tasks)

Pro tip: Enable DeepSearch for higher results. When tasks are complete, review notifications and refine prompts based on results quality.

PRESENTED BY IBM

🎙️Transforming HR with agentic AI

The Rundown: Inquisitive about agentic AI in HR? Steve Moss, Director of watsonx Americas at IBM, discusses HR, agentic AI, and the real-world impact of IBM’s transformation journey in the most recent episode of the AI in Motion podcast.

Dive into the main points of how IBM has:

• Used agentic AI and consolidated various HR tools right into a single interface

• Improved worker experience and operational efficiency

• Driven strategic shifts, boosted productivity, and scaled teams effectively

Click here to look at the video.

AI & NEUROTECHNOLOGY

🧠 LinkedIn co-founder bets on AI ultrasound helmet

The Rundown: LinkedIn co-founder and OpenAI investor Reid Hoffman just led a $12M funding round for Sanmai Technologies, which is developing AI-guided ultrasound devices for treating mental health conditions without surgery.

The small print:

• Sanmai’s consumer devices focus ultrasound waves on specific brain regions to treat anxiety, depression, and enhance cognitive function.

• The startup combines the ultrasound tech with AI coaching systems right into a helmet at a sub-$500 price point, targeting consumers’ in-home use.

• Hoffman joined Sanmai’s board through his Aphorism Foundation, saying non-invasive approaches are “much less dangerous” than tech like Neuralink.

• The corporate is currently testing anxiety treatments with a prototype at its Sunnyvale facility ahead of FDA trials.

Why it matters: Tech billionaires like Elon Musk, Jeff Bezos, Bill Gates, and now Reid Hoffman are all funding brain-tech startups. With an AI coach guiding treatments and a non-invasive approach, the beginning of the neurotech wave may find yourself being a lighter touch that is simpler to swallow for consumers than a full brain-computer interface.

QUICK HITS

🛠️ Trending AI Tools

• 💰 Taka – Your AI Personal CFO: Ask any personal finance query. Make confident decisions with a self-building canvas tailored to you*

• 🗣️ 11ai – ElevenLabs’ latest voice assistant with MCP integrations

• 💻 Mu – Microsoft’s fast, local AI for Windows Copilot + PCs

• 🎆 Imagen 4 Ultra – Google’s top image model, available via API & AI Studio

  *Sponsored Listing

💼 AI Job Opportunities

• 🔁 The Rundown – Strategic Partnerships (AI University)

• 🤝 OpenAI – Program Manager, M&A Recruiting

• 🖥️ Together AI – Senior Systems Administrator

• 📦 Groq – Senior Program Manager

📰 Every thing else in AI today

SimilarWeb data shows ChatGPT downloads on iOS hit 29M+ during the last 28 days, nearly surpassing downloads of TikTok, Facebook, and Instagram (33M) combined.

Sam Altman said that the ‘io’ lawsuit is “silly, disappointing and incorrect”, saying that founder Jason Rugolo made persistent attempts to get acquired by OpenAI.

Mira Murati’s Pondering Machines Lab is planning to develop custom AI models to assist businesses increase profits, in line with a brand new report from The Information.

Google released Gemini Robotics On-Device, a brand new VLA model that powers robotics dexterity and task completion without having an online connection.

Databricks & Perplexity co-founder Andy Konwinski launched the Laude Institute, pledging $100M to fast-track computer science breakthroughs for real-world impact.

XBOW revealed that its autonomous AI became the primary to surpass all humans on the HackerOne platform, also announcing a brand new $75M Series B funding round.

COMMUNITY

🎥 Join our next live workshop

Join our next workshop today at 3 PM EST with Max Brodeur-Urbas, Founder & CEO at Gumloop. By the top of the workshop, you’ll know easy methods to construct production-grade AI agents that mix context, reasoning, and actions with Gumloop.

RSVP here. Not a member? Join The Rundown University on a 14-day free trial.

See you soon,

Rowan, Joey, Zach, Alvaro, and Jason — The Rundown’s editorial team