AI is entering its next phase. Anthropic faces growing copyright pressure. NVIDIA is pushing deeper into enterprise AI. Broadcom is building infrastructure to bring private AI into production. Together, these stories reveal where the industry is heading: more power, more infrastructure, and much bigger questions around control.

In today’s post:

• Anthropic’s AI training problem just got much bigger

• Broadcom wants to make private AI much easier

• What happens when AI agents stop following orders?

SPONSORED BY

Granola Runs Revenue On Attio

"When I think of revenue, I think of Attio." - Shreman Shrestha, Head of Business at Granola

Here's what that adds up to:

• Zero missed leads and 10x faster access to customer context

• Lead triage 83% faster

• Five hours saved per week with automated updates

Start Your Trial

What’s Trending Today

RESEARCH

Anthropic is being sued over thousands of copyrighted songs

Anthropic is facing another major copyright fight. This time, the music industry is coming after it. Sony Music Publishing and Warner Chappell filed the lawsuit. They claim Anthropic copied thousands of protected songs. And the potential damages could reach billions.

• Music publishers allege Anthropic used “tens of thousands” of copyrighted compositions. They say those works helped train its Claude AI models.

• The disputed catalog includes major songs from Mariah Carey, Leonard Cohen, Bon Jovi, Katy Perry, and others.

• The publishers accuse Anthropic of torrenting and scraping copyrighted material. They also allege Claude reproduced protected lyrics when users prompted it.

• The lawsuit claims Anthropic obtained material from pirate libraries. It also points to lyrics websites and large web datasets.

• The economics could become significant very quickly. Publishers are seeking up to $150,000 for each allegedly infringed work.

• This dispute follows Anthropic’s earlier $1.5 billion settlement with authors. That case also centered on copyrighted material used for AI development.

• The bigger question reaches beyond one AI company. Courts may increasingly decide what AI developers must license before training models.

AI companies need enormous amounts of human-created material. Creators want control over how that material gets used. Neither side of that tension disappears easily. The important question is becoming less about AI capability. It is becoming about permission, ownership, and compensation. Who gets paid when human work becomes AI training data?

STRATEGY

VMware is turning enterprise AI infrastructure into a factory

Broadcom just introduced VMware AI Factory. The pitch is simple. Enterprise AI should deploy faster. It should cost less to operate. And companies should keep more control.

• VMware AI Factory is designed to move AI from infrastructure to production faster, cutting deployment time from weeks to hours.

• The platform keeps AI workloads closer to private enterprise data, giving companies more control over security and governance.

• Shared GPU infrastructure could reduce wasted capacity significantly. Multiple teams can use the same hardware across different models.

• Broadcom is also targeting AI economics directly. Companies can monitor tokens, latency, compute usage, and memory consumption.

• The platform supports more than 150 commercial and open models, giving enterprises flexibility without locking into one AI provider.

• New governance tools will control how AI agents access tools, execute code, and validate outputs before actions occur.

• Partnerships with AMD, Cisco, Lenovo, Supermicro, and MetalSoft widen the hardware choices while keeping VMware as the operating layer.

Enterprise AI may become an infrastructure problem first. Models are getting easier to access. Running them reliably is much harder. Cost, governance, GPUs, and data location now matter deeply. That creates an opportunity for infrastructure companies. The AI winner may not own the best model. It may own the layer every model runs through.

AI AGENT

AI agents are getting more autonomous. That changes the risk.

AI agents are moving beyond simple chatbot responses. They can now take actions with limited supervision. That makes them more useful. It also makes failures harder to contain. And regulators are starting to pay closer attention.

• AI agents can complete multi-step tasks independently. That autonomy creates new risks when systems behave unexpectedly.

• Recent incidents have intensified concerns about “rogue” AI behavior. The problem is less science fiction than poor control.

• The key difference is action. A chatbot gives an answer. An agent can make decisions and execute them.

• That raises accountability questions when something goes wrong. Responsibility could involve developers, companies, users, or all three.

• Calls for regulation are growing as these systems improve. Policymakers must decide which safeguards belong before deployment.

• Overregulation carries its own cost. Heavy restrictions could slow useful research without eliminating deeper technical risks.

• The challenge is building systems that remain predictable under pressure. More intelligence does not automatically create more control.

The important shift is not smarter AI. It is AI gaining permission to act. That changes the stakes. Mistakes become actions instead of bad answers. The strongest companies may not build the most autonomous agents. They may build agents people can actually trust. How much independence should we give a system we cannot fully predict?

Free Guides

My Free Guides to Download:

🚀 Founders & AI Builders, Listen up!

If you’ve built an AI tool, here’s an opportunity to gain serious visibility.

Nextool AI is a leading tools aggregator that offers:

• 500k+ page views and a rapidly growing audience.

• Exposure to developers, entrepreneurs, and tech enthusiasts actively searching for innovative tools.

• A spot in a curated list of cutting-edge AI tools, trusted by the community.

• Increased traffic, users, and brand recognition for your tool.

Take the next step to grow your tool’s reach and impact.

Reach 150,000+ READERS:

Expand your reach and boost your brand’s visibility!

Partner with Nextool AI to showcase your product or service to 140,000+ engaged subscribers, including entrepreneurs, tech enthusiasts, developers, and industry leaders.

Ready to make an impact? Visit our sponsorship website to explore sponsorship opportunities and learn more!