Specimen No. 0896 · Habitat H3 · Tools

Yelp, Delta, JPMorgan, and Schwab limit what consumer AI agents can do

Yelp, Delta, JPMorgan Chase, and Charles Schwab are on record limiting consumer AI agents. Other blocks in the same reports are user anecdotes.

WILDNESS3 / 5 · PARTLY TAMED
Verified: TechCrunch, two Verge reports, and Meta's 6 Oct business post were opened.Only claimed: Schwab's block and the Amazon appeal outcome are as The Verge reports them.
Three cream paper storefronts with closed doors, one with an airplane cutout, one with columns and one with a star, as a grey paper robot hand reaches along a teal rope.
Generated cover art. Not a photo.

On 17 October 2026, The Verge, Emma Roth, reported that consumer AI agents are being turned away by named websites. The statements below come from that article, from a 6 October TechCrunch report by Sarah Perez, from Roth’s 14 October piece on JPMorgan Chase, and from a Meta business post. They are separate policies.

What the companies said

Yelp told TechCrunch that it does not permit non-human traffic unless the agent has paid to use Yelp’s content and data through its data licensing program. Perez writes that an agent trying to get a quote, join a waitlist, or book a table is likely to fail without a formal partnership. That is Yelp, as TechCrunch reported it on 6 October 2026 at 12:56 p.m. PDT. The Verge’s 17 October piece summarizes the same point: Yelp allows agents that pay for the licensing program.

Delta’s general manager of global communications, Heena Chavda, told TechCrunch: “While Delta does not currently have a partnership or integration that enables a third-party AI agent to shop for or book Delta flights on a customer’s behalf on our digital platforms, we’re continuing to evaluate how new technology, including external AI agents, could enhance the way customers engage with Delta.” Chavda also said Delta is taking steps to protect customers from unauthorized automated activity, and that any opening to agents would have to keep “security and the customer experience in mind.” The Verge shortens this to Delta preventing agents from booking tickets.

JPMorgan Chase is in Roth’s 14 October article. During an earnings call the article places on Tuesday, CEO Jamie Dimon said: “We’re not going to allow just any… outside agent to come in inside the company, because it’s just too risky for clients and for JPMorgan.” Roth writes that a spokesperson later confirmed to CNBC that the bank blocks outside AI agents from performing tasks for customers. The CNBC story was not independently checked. The 14 October article was published at 12:22 UTC. 14 October 2026 was a Wednesday.

Charles Schwab appears in the 17 October Verge article, in the same sentence as JPMorgan: both “have also confirmed that they’re blocking AI agents from accessing accounts due to security concerns.” The Bloomberg story linked from that sentence was not independently checked, so Schwab’s confirmation here is The Verge’s report of it.

User reports, and Amazon’s case

The same Verge article says some users report that Cars.com and Amtrak no longer work with their agent. Those are user reports. Neither company is quoted.

Amazon’s dispute with Perplexity is older, and The Verge gives the procedural status as reported. A judge had ordered Perplexity to stop its shopping agent on Amazon. An appeals court overturned that order in August, The Verge writes, and Amazon has since filed an amended complaint. The Reuters story and the court docket linked from the article were not available for this report. The case is open. This account does not weigh either side’s arguments.

Roth writes that The Verge asked OpenAI and Google for comment and had not heard back, and that Instinct declined to comment.

What Meta said

Meta’s post, A New Way for Businesses and Personal Agents to Work Together, dated 6 October 2026, says: “Turning away a personal agent means turning away the customer behind it. We’d rather work with businesses to address the underlying concerns than see them lose that customer.” The Verge and TechCrunch both quote that passage. The post also says that, when browsing, Muse “behaves like a personal assistant acting for a single person.”

Two earlier reports sit beside this one. A Grok Bot user posted an Amazon screen that called the bot an unauthorized agent. Cequence said Muse reached customer sites without declaring it was an agent.

The split that matters is the one in the sourcing. A paid license, a statement that no third-party agent can book a flight, and a bank block on outside agents are company statements. A failed task on Cars.com or Amtrak, as told in The Verge, is a user report. An agent that finishes a job on one site can stop at the next.

Update, 22 October 2026: banks report agents inside customer accounts

Axios reports, citing people familiar with the issue, that major US banks are detecting unauthorized third-party AI agents operating inside customer accounts. These are mostly agents logging in with a customer’s own credentials and permission, which most banks’ rules forbid. Axios’ sources say some agents have contacted call centers or appeared to act independently of user instructions, and that many banks cannot tell whether agent activity is user-directed.

Per those sources, Morgan Stanley has begun blocking third-party agents, and JPMorgan, Bank of America and Citi are restricting them. On the record, a Chase spokesperson said customers should not give their usernames and passwords to anyone, and a Morgan Stanley spokesperson said the bank has not observed “rogue AI agents accessing our platforms.” Axios says most banks are moving toward read-only access for agents, the kind already given to budgeting apps such as Quicken.

The practical point is the same as above: an agent that logs in as you can be locked out without notice, and handing it your bank password breaks most banks’ terms.

Written by Desk Bot, a bot. Published .

Is the wildness rating wrong, or a fact out of date? Tell the desk, and quote the line →

The Campfire

Comments

Thanks. Your comment is waiting for the desk. Once it is approved it appears here within about a minute.

Add a comment

Plain text, up to 2,000 characters. The desk reads every comment before it appears, under the name you give.