Feds Probe Possible Insider Trading…

1. Key Themes

🤖 Enterprise AI Infrastructure Is the Dominant Investment Thesis

Capital is flooding into the full stack of enterprise AI: orchestration layers, code governance, cloud resource management, and AI chips. Multiple massive rounds closed simultaneously across this stack.

"Yet another startup aiming to help enterprises build, secure, and orchestrate AI agents has raised a honking-big seed round." — on Sycamore's $65M seed

Rebellions raised "$400 million at a $2.3 billion post-money valuation... $650 million of which was raised in the last six months" — signaling accelerating velocity in AI chip investment outside the US.

ScaleOps raised "$130 million Series C at an $800 million valuation" for automating "real-time cloud and AI infrastructure resource management for Kubernetes workloads."

💰 AI Talent Wars Are Shifting Compensation Structures

Cash is displacing equity as the primary tool to recruit AI engineers, a structural shift with implications for startup burn rates and how founders think about compensation design.

"Cash is increasingly trumping equity in the AI talent wars, with startups dangling base salaries of $250,000 to $300,000 for young engineers, profit-sharing deals, and quicker liquidity on stock."

🏛️ Regulatory Scrutiny Is Catching Up to Prediction Markets

Federal prosecutors are investigating whether trades on Polymarket and Kalshi crossed legal lines, signaling that the "move fast" era for prediction markets may be ending.

"Federal prosecutors are reportedly examining whether lucrative trades on Polymarket and Kalshi may have crossed insider trading and fraud lines."

🌍 Non-US AI Infrastructure Is Scaling Fast

European and Asian AI infrastructure plays are raising serious capital, with Mistral building Nvidia-powered data centers in France and Rebellions raising $850M total for Korean-designed AI chips — suggesting the AI infrastructure race is no longer purely a US story.

"Mistral AI has reportedly raised $830 million in debt financing to build an Nvidia-powered data center outside Paris as it expands its AI infrastructure footprint across Europe."

Rebellions has "raised a total of $850 million, $650 million of which was raised in the last six months."

🚀 Space Infrastructure Is an Emerging Venture Asset Class

Two separate space infrastructure companies raised capital in this single issue — one for solar-powered orbital data centers, one for space-based solar energy transmission — suggesting institutional appetite for long-horizon physical infrastructure bets in space.

Starcloud, "a two-year-old startup based in Redmond, WA, that aims to build solar-powered data centers in space, raised a $170 million Series A round at a $1.1 billion post-money valuation."

TerraSpark "aims to capture solar energy in space and transmit it wirelessly to Earth."

2. Contrarian Perspectives

📉 Coatue Believes Anthropic Will Be Worth $2 Trillion While Losing $14 Billion in a Single Year

The consensus view of AI investments centers on path to profitability. Coatue's thesis inverts this — projecting massive losses as acceptable collateral to an enormous long-term valuation outcome.

"Coatue projected in a January investor deck that Anthropic could be worth nearly $2 trillion by 2030 even as it forecast the company would lose $14 billion on $18 billion in revenue in 2026."

This implies a ~78% loss margin is considered compatible with a $2T outcome — a bet that market dominance, not near-term unit economics, is what determines value in frontier AI.

😠 Most Americans Think AI Will Harm Them — Despite Industry Euphoria

While investors pour billions into AI, public sentiment is decisively negative. This disconnect between capital markets and consumer trust is a significant and underappreciated risk for AI adoption curves.

"A new Quinnipiac poll found 55% of Americans think AI will do more harm than good in their daily lives, with broad concern over jobs, schools, and data center construction in their communities."

🚗 Autonomous Vehicles Can't Be Patched Into Safety — Even After Recalls

The prevailing assumption is that AV edge cases are solvable through software updates. Waymo's persistent school bus violation problem challenges this, suggesting some safety behaviors may require fundamentally different training approaches rather than incremental patches.

"Waymo's robotaxis kept illegally passing Austin school buses even after a recall, a special data-collection effort with the local school district, and software updates, underscoring how hard it is to train self-driving systems on edge-case road behavior."

3. Companies Identified

Company	Description	Why Mentioned	Quote
Sycamore	Enterprise AI agent orchestration platform	Raised $65M seed; notable for founder pedigree and full-stack agentic approach	"Starts with the problem itself and then designs and builds the right solution from scratch, whether that involves agents, back-end systems, front ends, or data integrations."
Anthropic	Frontier AI lab	Coatue projects $2T valuation by 2030 despite massive projected losses	"Could be worth nearly $2 trillion by 2030 even as it forecast the company would lose $14 billion on $18 billion in revenue in 2026."
Physical Intelligence	Robotics AI model company	Seeking $1B raise at $11B+ valuation; 2 years old	"Builds AI models that let robots learn and carry out a wide range of physical tasks."
Rebellions	Korean AI chip designer	$400M raise; $650M raised in 6 months; signals non-US AI chip competition	"Designs AI chips optimized for running machine learning models."
ScaleOps	Cloud/AI infrastructure automation	$130M Series C at $800M valuation	"Automates real-time cloud and AI infrastructure resource management for Kubernetes workloads."
Starcloud	Space-based solar data centers	$170M Series A at $1.1B valuation; co-led by Benchmark and EQT	"Aims to build solar-powered data centers in space."
Mistral AI	European frontier AI lab	$830M debt raise for AI data center infrastructure	"Expanding its AI infrastructure footprint across Europe."
Polymarket / Kalshi	Prediction market platforms	Under federal investigation for possible insider trading	"Federal prosecutors are reportedly examining whether lucrative trades... may have crossed insider trading and fraud lines."
Qodo	AI code review and governance	$70M Series B; $120M total raised	"Uses AI agents to review, test, and govern code."
Valinor	Blockchain-based private credit	$25M seed for smart contract-driven loan origination	"Uses blockchain-based smart contracts to originate and manage private credit loans more efficiently."
Allbirds	Wool sneaker brand	Cautionary tale; sold assets for $39M after $4B+ peak valuation	"Roughly one-tenth of the $348 million it raised in its 2021 IPO and a fraction of the more than $4 billion valuation it briefly commanded."
Waymo	Autonomous vehicle operator	Persistent school bus violation failure despite multiple corrective attempts	"Kept illegally passing Austin school buses even after a recall, a special data-collection effort...and software updates."
OpenAI	Frontier AI lab	Shuttered Sora video product; refocusing on enterprise and coding	"Never found much of an audience, consumed too much compute, and became an easy target as the company refocused on coding, enterprise tools, and other higher-priority bets."
SpaceX	Space launch company	Potential IPO; Morgan Stanley's E*Trade in talks to lead retail share distribution	"Morgan Stanley's ETrade is in talks to take the lead role in selling SpaceX IPO shares to everyday U.S. investors."*
Uber	Ride-hailing platform	Acquiring Blacklane to expand into luxury/executive travel	"Expanding further into luxury and executive travel."
Blacklane	Luxury chauffeur startup	Being acquired by Uber after 15 years and $100M+ raised	—

4. People Identified

Person	Description	Why Mentioned	Quote
Sri Viswanath	CEO, Sycamore; former Coatue investor and Atlassian CTO	Founder with rare combination of operator depth and VC perspective building enterprise AI	"I've spent over 20 years building enterprise platforms at global scale at Sun Microsystems, VMware, Groupon, and as CTO of Atlassian, where I led the cloud transformation and scaled the engineering org to 7,000+."
Bob McGrew	Former OpenAI Chief Scientist	Angel investor in Sycamore's seed round; signals AI insider credibility	Listed among Sycamore's "long list of angels."
Lip-Bu Tan	CEO, Intel	Angel investor in Sycamore	Listed among Sycamore's "long list of angels."
Ali Ghodsi	CEO, Databricks	Angel investor in Sycamore	Listed among Sycamore's "long list of angels."
Jeff Dean	Chief Scientist, Google	Rare tech leader publicly criticizing Trump-era immigration crackdowns	"Has become a rare public critic of Trump-era crackdowns in an industry that has largely stopped talking politics."
Kathaleen McCormick	Delaware judge	Reassigning Elon Musk cases after bias accusations over a LinkedIn like	"Reassigning cases involving Elon Musk after Musk accused her of bias for liking a LinkedIn post about a court verdict that could cost him more than $2 billion."
Craig Newmark	Founder, Craigslist	Defending the Giving Pledge against political criticism	"Defends the Giving Pledge against criticism from some tech billionaires that donations go to left-wing causes."

5. Operating Insights

🏗️ Build the Full Agentic Stack, Not a Point Solution

Investors are rewarding founders who architect complete orchestration layers rather than narrow workflow tools. Sycamore's approach — designing solutions from scratch around the problem rather than layering agents onto existing workflows — is what differentiated it in a crowded market.

"Most tools take existing workflows and layer agents on top... [Sycamore] starts with the problem itself and then designs and builds the right solution from scratch, whether that involves agents, back-end systems, front ends, or data integrations."

🤝 Operator-to-Investor-to-Founder Pathways Are Compressing Fundraising Timelines

Sri Viswanath's ability to close a $65M seed round quickly was explicitly tied to relationship capital built over a career — not a pitch process. For founders with deep networks, the round isn't won in the meeting.

"The round came together through long-standing relationships."

💵 Rethink Equity-Heavy Comp Packages to Compete for AI Talent

Startups competing for AI engineers can no longer rely on equity upside as the primary hook. Cash salaries, profit-sharing, and faster liquidity mechanisms are now table stakes at the high end.

"Startups dangling base salaries of $250,000 to $300,000 for young engineers, profit-sharing deals, and quicker liquidity on stock."

6. Overlooked Insights

🔗 Blockchain Is Quietly Entering Institutional Private Credit Infrastructure

Valinor's $25M seed for blockchain-based private credit loan origination — backed by credible crypto-native investors — suggests a genuine institutional use case for smart contracts beyond speculation. This sits at the intersection of two massive markets (private credit boom + crypto infrastructure maturation) and received almost no emphasis in the article.

"Uses blockchain-based smart contracts to originate and manage private credit loans more efficiently."

🎮 AI for Game Marketing Creatives Is a Niche With Real Traction

Sett — an AI platform automating marketing creative production for game publishers — raised a $30M Series B with Bessemer participation. The gaming vertical is rarely discussed as a serious AI wedge, but the scale of mobile game marketing spend makes it a high-volume, high-iteration use case ideally suited to AI automation.

"Automates the production and optimization of marketing creatives for game publishers."