Good morning. It's Wednesday, September 30, and we're covering Microsoft’s fight over enterprise AI data, Anthropic’s effort to shape AI ethics, and OpenAI’s push to bring work tools into ChatGPT.
Plus: yesterday’s poll results on AI regulation motives, an essay on how AI agents could disrupt advertising, and a new Real or AI challenge.
OpenAI introduced GPT-6.1 Sol as a lower-cost model with intelligence close to its flagship GPT-6 Astra. The company says Sol costs one-fifth as much as Astra, giving developers a cheaper option for demanding workloads.
OpenAI launched Space, a shared workspace where people and AI agents can work together, along with Pages for documents and a conversational tool for presentations. The release puts ChatGPT in more direct competition with Microsoft 365 and Google Workspace.
OpenAI is reportedly in talks to raise at least $30 billion at a roughly $1.4 trillion valuation in a pre-IPO round. The financing would come just months after its $122 billion March raise and could serve as a bridge to an eventual public offering.
A leaked Anthropic prospectus lists $4.6 billion in 2025 revenue and a $42 billion net loss, according to Fortune. Revenue rose 1,088%, while the filing warns investors that advanced AI could create catastrophic risks.
VIDEO
Sonnet 5.5’s Impressive 3D Builds
FORWARD FUTURE ORIGINAL
Why My Agent Punched the Monkey
This week, I have been thinking about the monkey because Amazon just threw Meta’s new Muse agent out of its store and blocked the bot from perusing the aisles or even crossing the virtual threshold. You remember the monkey, right?
It lived in a banner ad around the turn of the millennium and his job was to sit there looking punchable until you punched him. Punch the monkey and win! Perhaps an iPod. Perhaps $20. Perhaps absolutely nothing except the privilege of discovering that you had just volunteered your email address to seventeen mortgage brokers in Florida. → Read the full article here.
ENTERPRISE DATA
Microsoft Builds a Wall Around Business Data
Microsoft is opening a new front in the fight over business data. Its new IQ Sharing feature in Fabric lets companies share governed data and business context across organizational boundaries, while The Information reports that Microsoft is also trying to limit how rival AI systems reach data stored inside traditional enterprise apps.
The question is whether Microsoft can make sharing easier without turning its software into a toll booth for AI. Companies want agents to work across their data, but they also want control over who can see it, and which model can use it. Watch whether customers accept Microsoft’s rules or push for open connections that work with any AI provider. → Read the full article here. (Paywall)
MODEL VALUES
Who Gets to Write Claude’s Moral Code?
Anthropic uses a written constitution to shape how Claude behaves, including how the chatbot weighs safety, ethics, rules and requests from users. The company has also sought input from religious, philosophical and cultural communities as it studies how an AI system should handle difficult moral choices.
That work raises a basic governance question: when millions of people rely on one model, a private company’s values can influence decisions far beyond its own products. Readers should watch how much of that moral framework Anthropic makes public, whether customers can adjust it and who gets a voice when the rules change. → Read the full article here. (Paywall)
NEWS
What Else is Happening
Trump Pushes AI Without New Safeguards: The administration urged trust in AI’s promise after rogue agents probed government websites. (Paywall)
Samsung Invests $1B in Helix: Samsung and five affiliates are investing $1 billion in KKR-backed Helix, which builds AI data centers and power systems.
Meta Launches Muse for Small Business: Meta’s Muse analyzes sales, ads and social data for businesses, with paid plans starting at $20 monthly.
ChatGPT Plugins Get Apps: OpenAI added sidebar pages, interactive panels, custom file viewers and automated actions to ChatGPT plugins.
OpenAI Launches Always-On Avatar: OpenAI launched Dots, always-on personal agents powered by GPT-6 Astra for Pro and Business Premium.
AMD Buys World Labs for $8.2B: AMD agreed to acquire Fei-Fei Li’s World Labs in an $8.2 billion all-stock deal.
REAL OR AI
Can You Still Tell What’s Real?
POLL RESULTS
Sounding the Alarm or Building a Moat?
Here's how you voted: Self-interest led, with 46% seeing regulation as protecting major labs, while 30% cited both safety and strategy.
🟨🟨🟨⬜️⬜️⬜️ The risks are real, stronger safeguards are necessary (24%)
🟨🟨🟨🟨⬜️⬜️ Safety matters, but there’s a strategic advantage too (30%)
🟩🟩🟩🟩🟩🟩 Mostly self-interest — regulation protects the biggest labs (46%)
