// theme-commerce

All signals tagged with this topic

How Much Companies Actually Spend on AI

Azeem Azhar's team built the first deduped, bottom-up accounting of AI spending across the full stack. The methodology cuts through marketing claims and double-counted vendor revenue. For commerce, the granular view reveals where real money flows: infrastructure vs. applications vs. implementation services. It surfaces actual winners and losers in the AI economy rather than companies with the best PR. For trend watchers, this is the baseline data layer needed to distinguish which AI bets are attracting capital from which are running on hype.

Google's Epic Settlement Reshapes Mobile Payment Competition

Google's revised Play Store policies—including reduced commissions and alternative billing options—signal competitive pressure on Apple's 30% cut. Apple's commission erosion will come not from regulatory mandates but from Google demonstrating a viable alternative model, raising the cost of Apple maintaining current terms. Every percentage point Apple loses directly impacts services revenue, its highest-margin business segment.

Amazon seller exposes corrupt access scheme through chat app middlemen

An Amazon seller has documented how intermediaries operate on messaging platforms to connect merchants with Amazon employees who allegedly perform favors—account reinstatements, policy exceptions—in exchange for payment. The scheme exposes a vulnerability: suspended merchants are desperate enough to pay bribes, and employees have enough autonomy and weak oversight to monetize access. It bypasses Amazon's official appeals process and suggests the seller management infrastructure is both too rigid (forcing sellers to seek workarounds) and too porous (lacking audit trails on employee decisions).

Nvidia AI chip prices double on China black market under US sanctions

US export controls on advanced semiconductors have created a parallel market where Nvidia's flagship DGX B300 servers now trade at $1.1M—more than double retail—giving Chinese enterprises and state actors an expensive but available workaround to official restrictions. This arbitrage opportunity reveals the limits of unilateral export enforcement: sanctioned technology still reaches its highest-value buyers, but now with a 100%+ markup that effectively transfers wealth from Chinese purchasers to grey-market intermediaries rather than blocking access entirely. Chinese AI development isn't throttled by scarcity. It's throttled by cost, which money and state backing can solve.

The Profitability Question OpenAI Can't Outrun

OpenAI and Anthropic face a material barrier to IPO: neither has shown a path to sustained profitability at scale. Training and operating large language models demands capital intensity that keeps compounding. Billions in training costs, competitive pricing pressure, and unclear product-market fit beyond chatbots create a financial model public markets will scrutinize. Current business metrics cannot credibly answer the questions investors will ask. This is structural, not a timing problem better unit economics can fix. Market confidence in "AI profitability" remains fragile because the constraint is real.

Retail Media Networks Face Data Fragmentation Problem

Retail media networks have aggregated impressive ad inventory, but remain operationally fragmented. Amazon, Walmart, and Target each control their own data and refuse meaningful interoperability, forcing brands to rebuild campaigns separately. The economics only work if retailers can prove coordinated insights across channels improve performance—a requirement that demands breaking down the walled gardens protecting their highest-margin ad businesses. Without standardization, retail media risks becoming a channel tax rather than a data-enabled competitive advantage.

Algorithms Are Now Your Sales Reps—Commerce Strategy Must Adapt

As AI systems and LLMs increasingly handle product discovery and purchasing decisions, brands lose control of the customer journey to intermediaries they don't own or fully understand. Companies that treat algorithmic selling as a channel optimization problem rather than a structural change in how products reach buyers—requiring new approaches to product data, pricing transparency, and trust-building—will face compressed margins and vulnerability to platforms that own the algorithm-to-purchase pipeline.

Companies Turn to Prediction Markets for Business Risk Hedging

Kalshi's institutional volume is up 800% since November. Businesses are now using contract outcomes—election results, economic data, regulatory decisions—as hedging instruments rather than betting vehicles. Prediction markets are shifting from retail speculation into enterprise risk management infrastructure, much like commodity futures evolved from speculation into supply chain protection. The gap between traditional insurance pricing and real-time probabilistic pricing on Kalshi creates immediate pressure on legacy risk management vendors and opens a new asset class that regulators have spent years trying to constrain.

AI agents reshape Google Ads as autonomous shopping intensifies

Google Ads is moving from a display mechanism to an input layer for AI agents—where product feeds function as raw material for autonomous purchase decisions rather than human-browsable catalogs. This inverts the traditional search funnel: instead of ads pulling customers toward products, agents now pull products toward customers based on algorithmic preferences. Advertisers must optimize for machine readability and agent-friendly signals rather than human conversion psychology. The shift privileges scale, data quality, and technical feed infrastructure over creative messaging, rewiring which sellers win and lose in commerce.

Stripe Projects Targets AI Agents as Infrastructure Buyers

Stripe is repositioning its commerce infrastructure away from human-readable pricing pages toward machine-readable APIs designed for autonomous purchasing. The move reflects a shift in who's buying: AI agents, not humans, are becoming the primary buyers of cloud resources. This changes the go-to-market problem fundamentally. Vendors can no longer rely on sales friction, comparison shopping, or brand preference when agents execute purchases based on structured data, cost optimization algorithms, and programmatic contracts. The company that becomes the infrastructure layer for agent-to-vendor transactions gains significant leverage over which providers get selected and how pricing gets rationalized in an automated economy.

De Beers' blockchain gambit fails to stop lab-grown diamond surge

De Beers is deploying Tracr, its blockchain platform, to authenticate natural diamonds and create a provenance narrative—a defensive move that reveals the mined diamond cartel's real crisis: lab-grown stones are chemically identical, technically superior, and 40-50% cheaper, making authenticity claims irrelevant when consumers care more about price and sustainability. The 45% price collapse reflects a market that has already decided; blockchain traceability cannot rebuild demand for a product that younger buyers increasingly view as a commodity or ethical liability. De Beers is essentially paying to certify why its diamonds matter less, not more.

AI-Managed ETFs Are Beating Human Fund Managers

FINQ's algorithmic funds have posted better returns than traditional actively managed portfolios in early 2026. This is the first time AI asset management has moved from theoretical advantage to measurable outperformance. The $8+ trillion actively managed industry has long dismissed algorithmic competitors as unproven. Sustained performance gaps will accelerate capital flight toward lower-fee AI funds and force legacy wealth managers to either adopt similar technology or lose market share to startups. The contest is no longer AI versus humans—it's whether incumbents can transform faster than they're being disrupted.