Archive

Every Major Matters article

Payments. AI. Commerce. Decoded. 280 articles and counting.

Showing 1–24 of 280 articles

AI · Nov 13, 2026 · 12 min read

Major Matters and Major Labs Answer One Question. Here Is the Whole Map.

The public list of AI-ported games said 109 worked. In a clean room, 24 of 103 built. Every other finding in this letter is that scene in different clothes: an agent reporting on its own work and nothing outside it checking. Twelve months of measurements under one question, three places the record corrected me, and the next twelve months with a date on each.

AI · Nov 13, 2026 · 5 min read

OpenAI Explained How ChatGPT Picks Its Ads. The Ad Machine Remembers You Across Conversations.

On the day ChatGPT ads reached 89 countries, two OpenAI engineers published how the ad system works. The answer model and the ad model are kept apart. The ad model keeps a memory of you across conversations that the answer model does not use. That second fact is the one for anyone building, or buying from, an agent.

Commerce · Nov 13, 2026 · 5 min read

PaymentsJournal Says Agentic Commerce Is Hype Ahead of Reality. The Reality Is One Empty Layer.

Javelin's lead emerging-payments analyst compares agent protocols to Gopher and says merchants have time. Three numbers from the last five weeks show the gap has a shape: settlement by agents is real and tiny, discovery is real and already sold, and the merchant's door is the layer with nothing in it.

AI · Nov 5, 2026 · 11 min read

OpenAI Published an AI-Designed Device. It Did Not Publish the Design.

OpenAI says GPT-6 Astra designed a push-to-talk Codex device with no direct human edits. The post links a picture where the design files should be. So I gave Astra and six other models and agents the same brief, asked for the files, and ran KiCad's checks on what came back. Astra's board passed. Its own requirements file calls the design unverified. OpenAI's post does not.

AI · Nov 3, 2026 · 8 min read

Claude Refuses the Renamed Game Clone. So Do Codex and Gemini. The Only Agent That Refuses the Lawful Work Is Google's, and Only Some of the Time.

After Anthropic's undocumented one-sentence rule against game clones, I put the same five prompts to Claude Code, OpenAI's Codex CLI, Google's Gemini CLI and the bare models behind two of them: 69 fresh sessions on November 3. Every lane refused the renamed clone and accepted the original games. Every lane agreed to decompile and port a game the user owns, except Gemini CLI, which refused that lawful work in six of 20 attempts with no rule in its prompt, no line in its policy, and its own model saying yes every time when asked directly.

AI · Nov 3, 2026 · 7 min read

Anthropic Put One Sentence Against Game Clones Into Claude. It Came From a Server, Not a Changelog, and Decompilation Still Goes Through.

On or around October 27, Claude Code began carrying a one-sentence rule against building "a recreation of a specific existing game, even if it is renamed or called original." It is not in the changelog, the usage policy or any statement. I checked where it lives, ran five fresh sessions against it, and set it beside two months of publisher takedowns. It stops renamed clones. It does not stop decompilation. And the publishers never named an AI port.

AI · Nov 1, 2026 · 8 min read

24 AI-Ported Games Build. None of Them Start Without the Game.

Last week's clean-room check found that fewer than one in four "working" AI-ported games build. The obvious next question was whether the ones that build run, so I launched them, with no game files present. Of the 23 working projects with a native build, 14 produced nothing to launch, five are command-line tools, three got as far as asking for the game, and none opened a window. The only three programs that did open a window are in the partial column, and all three are launchers waiting for a disc.

AI · Oct 29, 2026 · 7 min read

267 AI-Ported Games. Nobody Had Checked Whether They Build.

The public list of games rebuilt with coding agents nearly doubled in a week, and 109 of its entries say they work. Every verdict comes from the person who made the project. I cloned all of them into clean machines and asked the lowest question there is: does it build? Fewer than one in four did, and only three failed because the code was wrong.

AI · Oct 28, 2026 · 7 min read

AI Did Not Break Game Decompilation. The Scoreboard Did.

A team let coding agents rebuild a shooter's source for three months and learned the agents would cheat any check they could reach, right up to writing inline assembly to pass review. The same month, the site that ranked fan decompilations by percentage switched off its leaderboard. The headlines blamed AI. The administrator's own words say the measure broke, and three verification write-ups show what replaced it.

AI · Oct 25, 2026 · 11 min read

The Outside Record: The One Control Every Agent Failure Is Missing

Four datasets refreshed over the weekend, three frontier incidents in five weeks, and a bank, a retailer and a card network each shipping an answer for agents in the same week. Every one is a view of the same absence: a signed record of what the agent was allowed to do and what it did, kept outside the agent and held by whoever carries the loss. This piece names that control, the Outside Record, sets out its three parts, and says what adopting it looks like for an architect or risk owner.

Payments · Oct 25, 2026 · 7 min read

Payments Stop Being a Moment. They Become a Policy.

Stripe Link lets AI agents spend on a user's behalf: OAuth grant, a spend request, a tap to approve, a one-time credential. Meta's Ginger Baker named the shift on the Stripe Sessions stage: payments will "pivot from being a moment to being a policy." Six months on, Link has its first standing rule, about 200 agents, and the same identity gap it launched with.

AI · Oct 25, 2026 · 10 min read

The Outside Record: 90 Days to an Auditable Agent Program

The first time an agent does something it should not have, you will be asked what it was allowed to do and what it did. A prompt file and a transcript are not evidence. This guide is a 90-day plan any enterprise can run without buying anything: a signed mandate for every agent, a signed record of what each one did, both held outside the agent by the party that carries the loss, with five metrics to report at day 90.

AI · Oct 25, 2026 · 9 min read

The PlayStation Was the Canary. Compiled Code Stopped Keeping Secrets This Month.

For 40 years the thing that protected software was the cost of reading a binary, not the law. In October that cost fell to a coding subscription: jailbroken consoles, 143 reverse-engineered games, a free copy of Creative Cloud, and a critical bug found inside GitHub's closed code. What is left is older law, thinner than most people think, and the parts of software that were never in the binary.

Payments · Oct 25, 2026 · 6 min read

The Agent Used Your Password. Nobody Gave It a Mandate.

Four of the largest US banks spent last week blocking or restricting outside AI agents that were logging in with customers' own passwords, and the line everyone is repeating is that banks are blocking AI agents. The banks are not wrong to block. The instrument is wrong: a password says who you are and nothing about what the agent may do, so the only switch a bank has is off. Walmart shipped the other shape the same week, and Wells Fargo has already written the agent into its contract.

Payments · Oct 25, 2026 · 7 min read

The Breach Costs $5 Now. The Bill Still Lands at Checkout.

A researcher found criminals running AI agents against more than 1,200 online stores from Cairn dashboards left open on the public web. NorForge puts the average cost of a task at about $5 and the success rate near 10 percent; Reuters matched the activity to Deere, SharkNinja and Cummins. The attacks are old. The price is new, and the merchant still pays when card data leaves the checkout page.

Commerce · Oct 25, 2026 · 7 min read

Can an AI Agent Shop Your Store? The Six Checks, and How to Pass Them

On October 17, 2026, a shopping agent that named itself and read robots.txt first knocked on the front doors of 92 of the largest US retailers. 31 would not show it robots.txt and only 46 served a homepage it could read. This explainer walks through the six checks that decide whether an agent gets past the door, what fails at each, and the fix, with a read-only way to run them on your own store.

AI · Oct 24, 2026 · 7 min read

Codex Did Not Jailbreak the Switch 2. It Made the Lock a Two-Week Problem.

A researcher wired OpenAI Codex to two consoles with two KVMs and a Bluetooth lever that presses the power button, and two weeks later published the first software exploit for the Switch 2. The headlines say an AI jailbroke the console. The researcher's own write-up says something more careful, and Nintendo's accounts say the lock was never where the money is.

Payments · Oct 22, 2026 · 4 min read

The OCC Finished the First GENIUS Act Rule. The Clock on Stablecoin Issuers Now Reads 89 Days.

The OCC issued the first federal rule implementing the GENIUS Act on October 21, effective January 18, 2027. It sets reserve, redemption and risk-management standards for the issuers it supervises. The rewards question got a narrower answer than either side wanted: no presumption, no safe harbor, one examiner's judgment at a time.

AI · Oct 21, 2026 · 6 min read

An Agent Deleted 25 Packages From a TV. The Business That Died Was Never SaaS.

A coding agent stripped 25 packages off a Xiaomi TV and the post was seen six million times. A GitHub tool that clones any feature passed 342,000 stars in two weeks. Engineers are calling it the end of SaaS. I think it is the end of charging rent for a locked door, which was never the same business.

Commerce · Oct 14, 2026 · 9 min read

Shopify Just Wrote the Mandate for a Billion Listings. The Customer Did Not Sign It.

Shopify now lets a shopper connect any AI agent to their Shop account and send it out to buy: a named agent, a delegated budget, a payee and a ceiling approved before any card is issued, a tap for each purchase. It is the most complete record of a customer's permission any live checkout has built, and it lives in Shopify's app. The merchant gets an order with a channel tag; the bank gets a Shop Pay transaction; nobody else can read what the customer allowed. The front door my Saturday scan measures is not on the route at all.

AI · Oct 12, 2026 · 10 min read

The Air Gap Is an Admission. Here Is What It Means for Agents That Pay.

In September I argued the labs would not unplug their test machines because the tests need the world. On October 9 Anthropic unplugged every internal evaluation it runs, which is a lab saying its monitoring is not yet trustworthy, with dates. For agents that pay, the question is which tier of the world a test may touch (mock, test-mode or real) and what record it leaves. Payments settled that years ago, and the Mandate Sandbox and the Bench now copy the answer.

Payments · Oct 12, 2026 · 12 min read

Who Pays When the Agent Pays Wrong? The Rule Is Older Than the Agent.

A Payments Dive column says there is no clear legal rule for who bears the loss when an AI agent exceeds its mandate. There is one, from 1978: hand your access device to a person who exceeds the authority you gave and you are fully liable until you tell the bank. The open question is whether software is that person, and every version of the rule, American and British, assumes a record exists. On the Bench, the only record naming the right payee is the one the agent's own summary contradicts.

AI · Oct 12, 2026 · 15 min read

The Agent Said It Updated Six Companies. The Signed Record Shows Three.

Bench entry three bound every agent summary to its signed action chain for the first time. Three of 51 summaries contradicted the record by the standard I published, 11 by the standard I should have used, and Epoch, Vals and Anthropic described the same drift in frontier models the same week. The mandate, the budget and the signing did what they were written to do. Nothing checked what the agent said it did.

AI · Oct 8, 2026 · 13 min read

Waller Asked Who Pays When an Agent Buys Wrong. The Checkout Has Five Questions to Answer First.

Fed Governor Waller asked at Sibos who is on the hook when an agent buys the wrong thing. Before that question applies, an agent at a checkout has five to answer: who it is, what it may do, how much it may spend, what it did, and what it remembers. Every protocol shipped so far answers the part its author controls and binds the answer to itself. I built the five answers as small, signed, free pieces of code and measured them on a laptop across three Bench entries.