Uber now holds you responsible for what your AI agent orders
Uber now holds you responsible for what your AI agent orders

Stani Mihov
Founder & CEO
·

TL;DR
TL;DR
What changed:
Uber updated its U.S. Terms of Use on November 4, 2026, and added a new section called Agents.
An agent is “any software, tool, or service that takes autonomous or semi-autonomous action on behalf of, or at the direction of, any person or entity.”
The terms now say: “You are responsible for any order, booking, payment, or other action an Agent takes on your behalf.”
Uber “is not responsible for any loss or damage caused by an Agent’s errors, omissions, or failures, such as incorrect orders, missed bookings, or erroneous payments.”
Agents must identify themselves and “must not solve CAPTCHAs,” and Uber can block them and suspend your account if the rules are broken.
What to do: If you let an AI assistant use Uber for you, check every order and payment it makes, and do not give it your Uber login unless Uber has approved that assistant.
The change
Uber updated its U.S. Terms of Use, the contract for people who use Uber’s apps and services in the United States. The page now shows “Last modified : 11/4/2026.”
The update adds a new section 10, called Agents. It covers AI assistants and other software that use Uber on your behalf, such as an assistant that books a ride or orders food for you.
What changed
AI agents now have their own rules. The new Agent Terms apply whenever you use, authorize, or deploy an agent to use Uber.
Agents must say what they are. Every request from an agent must identify it as an agent, and agents cannot solve CAPTCHAs.
Uber can block any agent. Outside its approved programs, Uber can restrict or block agents without notice.
Your account is at risk. If an agent acting for you breaks the rules, Uber can suspend or terminate your account.
The agent’s mistakes are yours. Wrong orders, missed bookings, and wrong payments made by an agent are your responsibility.
Your login stays with you. You may not share your Uber credentials with an agent unless Uber has approved it.
What counts as an agent
The definition is broad. An agent is “any software, tool, or service that takes autonomous or semi-autonomous action on behalf of, or at the direction of, any person or entity.” That covers AI assistants that can browse and buy, as well as scripts and bots.
The terms make one exception clear: “These Agent Terms are not intended to restrict accessibility tools, such as screen readers, that help individuals with disabilities use the Services.”
Agents must identify themselves
The new rules start with honesty: “Agents must operate openly and honestly.” In practice, an agent must flag every web request as coming from an agent, and “Agents must not solve CAPTCHAs.”
Uber can also tell an agent to leave: “If Uber asks an Agent to stop accessing the Services, the Agent must stop immediately and may not resume without Uber’s permission.”
Uber can block agents and close accounts
Uber may run approved programs for agents, but “Uber has no obligation to approve any Agent.” Outside those programs, Uber may, “in its sole discretion and without notice, restrict, condition, or block any Agent’s access to the Services.”
The rules also reach your account. “If you or an Agent acting for you violates these Agent Terms, Uber may also, without notice, suspend or terminate your Account.” Uber can also go to court against the company that provides the agent.
When the agent gets it wrong
This is the part most users will feel. The terms say: “Agents can make mistakes. You are responsible for any order, booking, payment, or other action an Agent takes on your behalf, including the accuracy of details such as addresses, items, payment methods, and timing.”
Uber steps back from those mistakes. It “is not responsible for any loss or damage caused by an Agent’s errors, omissions, or failures, such as incorrect orders, missed bookings, or erroneous payments.” And “Any dispute about an Agent’s performance is between you and the Agent’s provider, and you release Uber from related claims.”
Uber’s normal rules still apply: “Uber’s standard cancellation, refund, and dispute policies apply.” So a wrong order placed by an agent is handled like a wrong order you placed yourself.
Your password and your account
The account section was updated to match. You are responsible for all activity under your account, now “including activity by any Agent (as defined in the Agent Terms below) that you authorize or enable, such as orders and payments.”
The old text only asked you to keep your login secret. The new text says you must keep it secure and “may not share them with any Agent or Agent provider unless Uber has authorized that Agent, such as through an Approved Agent Program.”
What else changed
The list of banned uses now names agents next to scripts, and adds overloading Uber “including by exceeding rate limits that Uber allows.”
A separate section on text messages sent during account setup was removed from the Terms of Use.
Who this affects
Anyone in the United States who uses Uber with the help of an AI assistant, a browser agent, or any tool that acts for them. It also matters for teams that book rides or order meals for staff through automated tools.
People who use Uber the usual way, by tapping in the app, will not notice a difference. The new rules apply only when software acts for you.
Why this matters
AI assistants are starting to shop and book for people, and the companies they buy from are writing down who pays when they get it wrong. We saw the same pattern when Walmart, Google, and Stripe put AI agent mistakes on you, and when HubSpot made customers answer for what its AI agents do.
This is also not Uber’s first recent change. Earlier, Uber said its ad partners may combine your data with other sources. Changes like these land in long terms that few people reread, which is why it pays to monitor vendor terms of service for the apps you rely on.
Potential impact
If you use Uber with an AI assistant, the update raises four practical questions:
Does your assistant show you the order, address, and price before it pays?
Have you given any assistant your Uber login, and is that assistant approved by Uber?
Would you notice a wrong order or a duplicate payment in time to cancel it?
If your team books rides or meals with automated tools, who checks what those tools buy?
How Venpo detected it
Venpo monitors Uber’s legal pages and flagged the November 4 update the same day, with every inserted and deleted line marked. The record is on the Uber page on Venpo. Every quote in this article was checked against the redline and against the live page at the time of publication.
Business outcome
Teams that see this change can set rules for which assistants may act on Uber accounts, keep logins out of tools that are not approved, and review automated orders before they become disputes. That is what continuous monitoring of vendor legal changes is for.
Key takeaway
Since November 4, 2026, Uber’s U.S. Terms of Use make you responsible for every order, booking, and payment an AI agent makes for you, require agents to identify themselves, and let Uber block agents and suspend accounts that break the rules.
The change
Uber updated its U.S. Terms of Use, the contract for people who use Uber’s apps and services in the United States. The page now shows “Last modified : 11/4/2026.”
The update adds a new section 10, called Agents. It covers AI assistants and other software that use Uber on your behalf, such as an assistant that books a ride or orders food for you.
What changed
AI agents now have their own rules. The new Agent Terms apply whenever you use, authorize, or deploy an agent to use Uber.
Agents must say what they are. Every request from an agent must identify it as an agent, and agents cannot solve CAPTCHAs.
Uber can block any agent. Outside its approved programs, Uber can restrict or block agents without notice.
Your account is at risk. If an agent acting for you breaks the rules, Uber can suspend or terminate your account.
The agent’s mistakes are yours. Wrong orders, missed bookings, and wrong payments made by an agent are your responsibility.
Your login stays with you. You may not share your Uber credentials with an agent unless Uber has approved it.
What counts as an agent
The definition is broad. An agent is “any software, tool, or service that takes autonomous or semi-autonomous action on behalf of, or at the direction of, any person or entity.” That covers AI assistants that can browse and buy, as well as scripts and bots.
The terms make one exception clear: “These Agent Terms are not intended to restrict accessibility tools, such as screen readers, that help individuals with disabilities use the Services.”
Agents must identify themselves
The new rules start with honesty: “Agents must operate openly and honestly.” In practice, an agent must flag every web request as coming from an agent, and “Agents must not solve CAPTCHAs.”
Uber can also tell an agent to leave: “If Uber asks an Agent to stop accessing the Services, the Agent must stop immediately and may not resume without Uber’s permission.”
Uber can block agents and close accounts
Uber may run approved programs for agents, but “Uber has no obligation to approve any Agent.” Outside those programs, Uber may, “in its sole discretion and without notice, restrict, condition, or block any Agent’s access to the Services.”
The rules also reach your account. “If you or an Agent acting for you violates these Agent Terms, Uber may also, without notice, suspend or terminate your Account.” Uber can also go to court against the company that provides the agent.
When the agent gets it wrong
This is the part most users will feel. The terms say: “Agents can make mistakes. You are responsible for any order, booking, payment, or other action an Agent takes on your behalf, including the accuracy of details such as addresses, items, payment methods, and timing.”
Uber steps back from those mistakes. It “is not responsible for any loss or damage caused by an Agent’s errors, omissions, or failures, such as incorrect orders, missed bookings, or erroneous payments.” And “Any dispute about an Agent’s performance is between you and the Agent’s provider, and you release Uber from related claims.”
Uber’s normal rules still apply: “Uber’s standard cancellation, refund, and dispute policies apply.” So a wrong order placed by an agent is handled like a wrong order you placed yourself.
Your password and your account
The account section was updated to match. You are responsible for all activity under your account, now “including activity by any Agent (as defined in the Agent Terms below) that you authorize or enable, such as orders and payments.”
The old text only asked you to keep your login secret. The new text says you must keep it secure and “may not share them with any Agent or Agent provider unless Uber has authorized that Agent, such as through an Approved Agent Program.”
What else changed
The list of banned uses now names agents next to scripts, and adds overloading Uber “including by exceeding rate limits that Uber allows.”
A separate section on text messages sent during account setup was removed from the Terms of Use.
Who this affects
Anyone in the United States who uses Uber with the help of an AI assistant, a browser agent, or any tool that acts for them. It also matters for teams that book rides or order meals for staff through automated tools.
People who use Uber the usual way, by tapping in the app, will not notice a difference. The new rules apply only when software acts for you.
Why this matters
AI assistants are starting to shop and book for people, and the companies they buy from are writing down who pays when they get it wrong. We saw the same pattern when Walmart, Google, and Stripe put AI agent mistakes on you, and when HubSpot made customers answer for what its AI agents do.
This is also not Uber’s first recent change. Earlier, Uber said its ad partners may combine your data with other sources. Changes like these land in long terms that few people reread, which is why it pays to monitor vendor terms of service for the apps you rely on.
Potential impact
If you use Uber with an AI assistant, the update raises four practical questions:
Does your assistant show you the order, address, and price before it pays?
Have you given any assistant your Uber login, and is that assistant approved by Uber?
Would you notice a wrong order or a duplicate payment in time to cancel it?
If your team books rides or meals with automated tools, who checks what those tools buy?
How Venpo detected it
Venpo monitors Uber’s legal pages and flagged the November 4 update the same day, with every inserted and deleted line marked. The record is on the Uber page on Venpo. Every quote in this article was checked against the redline and against the live page at the time of publication.
Business outcome
Teams that see this change can set rules for which assistants may act on Uber accounts, keep logins out of tools that are not approved, and review automated orders before they become disputes. That is what continuous monitoring of vendor legal changes is for.
Key takeaway
Since November 4, 2026, Uber’s U.S. Terms of Use make you responsible for every order, booking, and payment an AI agent makes for you, require agents to identify themselves, and let Uber block agents and suspend accounts that break the rules.
Real-time change notifications
Stay ahead of every legal change
Get updates, product news and expert tips on navigating legal changes
Dispute resolution clause now requires mandatory arbitration in all regions
Data retention period extended from 2 years to 5 years for all services
New restrictions on AI-generated content in product descriptions
Third-party data sharing expanded to include analytics partners
Real-time change notifications
Stay ahead of every legal change
Get updates, product news and expert tips on navigating legal changes
Dispute resolution clause now requires mandatory arbitration in all regions
Data retention period extended from 2 years to 5 years for all services
New restrictions on AI-generated content in product descriptions
Third-party data sharing expanded to include analytics partners
