How to Beat Frozen Merchant Accounts
The hold, reserve, and freeze clauses that lock up merchant money at Stripe, Square, PayPal, and Cash App, and the leverage that gets it back.
Broadcasting from the fine print
A working lawyer's desk, on the air.
Featured
Two long form specials for the weekend, picked from the desk.
The hold, reserve, and freeze clauses that lock up merchant money at Stripe, Square, PayPal, and Cash App, and the leverage that gets it back.
The human authorship line, what it does to your logo, your codebase, and your ad campaign, and the contract language that limits the damage.
Long forms produced with Google NotebookLM from the stations' source reviews; two AI hosts.
Program guide
4 stations · 36 episodes · 5.5 hours of programming
Pseudo-live: everyone who tunes in hears the same moment. Press a preset on the radio to join the broadcast in progress, or open a station drawer below to play any episode from the top, jump to a chapter, or read along.
Terms of service and privacy policies, read out loud and graded. I read the fine print so you don't have to. (12 episodes, 2:03:23 of programming.)
Cash App, Square, PayPal, Stripe, and Anthropic all updated their fine print this summer. The grades are in, and nobody made the honor roll.
Holds, limitations, and reserves in the terms of Stripe, Square, PayPal, and Cash App, and what each grip actually does to your cash flow.
Vendor by vendor, what the AI industry's own published terms say happens to your prompts, files, outputs, and feedback after you press enter.
The unilateral amendment clause under the magnifying glass: five moving parts that let the other side rewrite the deal after you agreed to it.
The password is right and the account is gone. What the fine print says about suspension, termination, and your data, and how to be ready first.
The clause that says the software running your business owes you, at most, the price of an office chair when everything goes wrong. Read slowly.
Renewal defaults, then-current rates, and plan changes that arrive with the notice period of a sneeze. The three gears of the subscription machine.
You usually keep ownership of your uploads, and it usually matters less than you think. License grants, feedback clauses, and improvement rights.
Binding arbitration, class action waivers, opt-out windows, and the contract deadlines that can run out before your legal ones do.
Affiliates, advertisers, legal demands, and the sale of the company itself: the four doors your data, and your customers' data, walks out of.
The recurring rundown of the most consequential recent terms changes on the watchlist: who changed what, and where the scores stand now.
A two-host AI roundtable over the station's source reviews of Cash App, Square, PayPal, Stripe, Anthropic, Kraken, OpenAI, and Gemini, including the hold, reserve, and freeze clauses that lock up merchant money. An hour-scale weekend listen.
Produced with Google NotebookLM from the station's source reviews; two AI hosts.What crossed a working business lawyer's desk. (11 episodes, 1:19:04 of programming.)
PROGRAM · WASHINGTON TO CONTRACT
Three stories: the government switched off the most powerful AI model for nineteen days, the White House joined the release pipeline, and the biggest crypto bill missed its own party.
An AI model went dark for nineteen days on a government order. The contract clauses that decide whether the next outage is an inconvenience or a crisis.
Washington and the states both claim the right to regulate AI. Whose rule ends up in your contract, and the compliance allocation clause that decides.
The crypto market structure bill is stalled, and the three fights blocking it land on ordinary ledgers: payments, custody, treasury, and counterparties.
The customs bill goes to the importer of record, but the economic cost goes wherever the contract sends it. The five clauses that decide who pays.
United States export law can reach your customer list, your access controls, and your uptime. What the law reaches, who you let in, where you serve.
The noncompete is the least reliable tool in the drawer. What still works, what fails, and how to build a restrictive covenant stack that survives.
Two lists: what changes your operations now under law already in force, and what you monitor with a calendar entry instead of a budget line.
A procurement standard written for federal contractors ends up, four links later, in the vendor agreement of a company that never sold to a government.
Not an immigration law episode. What corporate documents, contracts, and operating plans should already say before a key person's status changes.
The quarterly flagship: nine items, three lists. What to act on now, what to calendar, and what is drawing more attention than it deserves.
One case file per episode: the evidence, the number, the deadline, the morning after. Noir for the fine print. (6 episodes, 42:39 of programming.)
Does a demand letter actually do anything? The honest answer, the anatomy of the letters that work, and the mistakes that quietly kill the rest.
The strength of a demand is decided before the drafting starts. What I ask every client for, and why each item in the packet changes the letter.
No sentence in the letter gets written wrong more often than the demand line. Four numbers that go into one, and the order in which to compute them.
The deadline is the most misunderstood move on the board. People think it pressures the recipient. Mostly, it commits the sender.
The letter went out, the deadline passed. The answer takes one of four shapes: silence, the lawyer letter, the lowball, and capitulation.
The file where a lawyer talks himself out of a fee: when a demand letter is the wrong way to spend your money, and what to do instead.
The terms of the machines: who owns the output, who trains on your data, who pays when it is wrong. (7 episodes, 1:23:39 of programming.)
The question from almost every consult now: who owns what your AI makes, and who pays when it gets something wrong? Ownership first.
Whether your logo, your codebase, or your ad campaign is an asset you can defend or just a file on a server. The human authorship line, all the way down.
Every vendor now recites the same sentence: we do not train on your data. Why that sentence alone says almost nothing, and the four definitions underneath.
Before your business signs up for any AI tool, seven clauses decide nearly all the legal risk. What good looks like, plus one test question for each.
The internal AI policy your business actually needs fits on one page, because a policy has to work at 4:45 on a Thursday, not in the annual training.
The oldest question in the law, wearing new clothes. The disclaimer stack and its limits, and where liability lands when someone relies on bad output.
A two-host AI roundtable on AI output ownership, the human authorship line, and what an AI-native solo practice actually looks like, produced from the station's source materials. A weekend listen just under the hour.
Produced with Google NotebookLM from the station's source reviews; two AI hosts.Every episode is written and legally reviewed at my desk before it airs. AI voice technology produces the narration and keeps the station running around the clock; the judgment about what is worth saying is mine. The source material behind each episode is linked from the transcripts and from the ToS Watchdog and Privacy Watchdog libraries.