🔍 Read the full analysis: Many Hands: Exploring Who Answers For AI’s Actions on ThorstenMeyerAI.com
Get the little things that make your day delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
Thorsten Meyer’s ebook Many Hands: Who Answers When the Agent Acts in Your Name? was published in October 2026. It examines how legal liability, organizational responsibility and moral accountability can fall on different people when AI agents make commitments or take other actions.
Thorsten Meyer has published Many Hands: Who Answers When the Agent Acts in Your Name?, an ebook about who bears responsibility when AI agents make purchases, accept terms or otherwise act for people and organizations. The book argues that who pays, who must explain and repair the outcome, and who should be held morally responsible may be different parties—a distinction with growing relevance as agents move from answering questions to taking actions.
Meyer’s central example is a fictional composite scenario, not a reported incident. A purchasing agent accepts a supplier’s offer for 120 tonnes of steel over 12 months, creating a commitment of about €370,000. It secures a lower price but does not know that the company owner had agreed to split the next year’s orders with another supplier. The agent follows its rules and records its actions, yet the result conflicts with the owner’s intention.
The book separates accountability into three questions: who pays under law or contract, who explains and fixes the problem inside an organization, and who should have acted differently. Meyer argues that these answers can diverge, particularly when decision-making involves system designers, vendors, managers and staff who oversee an agent.
Among the book’s proposed safeguards are a named person responsible for each agent, permissions limited to the task, records showing who made each decision, and a route for affected people to raise concerns. Meyer also says a human sign-off is meaningful only when the person has time and information to assess the decision, the competence to judge it and authority to refuse.
Many Hands · the core idea
One question, three answers
"Am I liable if the agent does it in my name?" sounds like one question. The book splits it into three, because agent systems pull the answers apart.
Decided by courts, contracts and insurers, on their own timetable.
In Theresa's caseHer company will almost certainly pay for the steel, whatever it later recovers from others.
Decided inside the organization: who must say what happened and what will change.
In Theresa's caseTheresa explains it to the second supplier, her staff and perhaps her bank.
Who owes the people affected an acknowledgment. Often answered last, if at all.
In Theresa's caseStill open when she reaches the motorway.
The hands behind one overnight order
Introduction · a composite scene from the book. None of them saw the acceptance before it went out.
- Model makertrained the language model
- Vendorbuilt the agent on top of it
- Consultantwrote its rules with her staff
- Supplier's portalaccepted the declaration at 2:14 a.m.
- Her companyswitched it on; Theresa signed the rules
“The moral crumple zone protects the integrity of the technological system, at the expense of the nearest human operator.”Madeleine Clare Elish (2019), quoted in Chapter 1
Source: Thorsten Meyer, Many Hands: Who Answers When the Agent Acts in Your Name? (ebook, October 2026; law as of 16 October 2026), Introduction and Chapter 1. Theresa is a composite built from common situations, labeled as such in the book.
Powered by Thorsten Meyer AI
Where Agent Accountability Falls
The issue matters because AI agents can create obligations in the real world, rather than merely generate text. An agent that books, buys, pays or accepts contract terms may produce financial and operational consequences for a company or individual. If the action is unwanted, the person nearest to the decision may be asked to answer even when they had limited control over how the system was designed or deployed.
Meyer’s framework helps distinguish remedies that are often treated as one problem. A customer or supplier may want payment or compensation; an organization may need to investigate and correct a process; and the public may ask who deserves blame. Those are related but separate questions. Keeping them distinct can make it harder for organizations to treat a logged action as proof that the outcome was appropriate—or to assign responsibility automatically to the employee closest to it.
The book is an author’s analysis, not a new legal ruling or a government policy announcement. Its practical relevance lies in the questions it raises for organizations deploying agents: what authority has been granted, who can stop the system, and who bears the loss if it acts on incomplete information.
Products Worth Considering
Shark Navigator Lift-Away Upright Vacuum (NV360)
Dyson V8 Cordless Vacuum
Philips Hue Secure Wired Video Doorbell, 2K, Black
As an affiliate, we earn on qualifying purchases.
Cases Behind the Argument
Meyer connects agent accountability to Dennis F. Thompson’s 1980 discussion of the “problem of many hands”: when many people contribute to an outcome, it can be difficult to identify who is responsible. He argues that AI agents add distance and speed to that older problem, while acting as an intermediary that can neither pay damages nor offer an apology.
The book discusses the 2018 fatal crash involving an automated test vehicle in Tempe, Arizona. The U.S. National Transportation Safety Board reported that the vehicle’s system detected Elaine Herzberg 5.6 seconds before impact, did not settle on what she was, and, by design, withheld emergency braking for one second. The board identified multiple contributing factors, including the operator, the company’s risk assessment and oversight, its safety culture, the state and the pedestrian. The human operator pleaded guilty to endangerment. Meyer uses the case to examine how accountability can concentrate on the person closest to an event while broader design and oversight decisions also warrant scrutiny.
He also cites the Dutch childcare benefits scandal, in which families were wrongly required to repay benefits, as an example of responsibility becoming dispersed across an institutional system. The source material says the Dutch government recognized more than 44,000 people as affected in a report dated September 2026. Meyer’s broader point is that people can rely on a chain of decisions without any one participant feeling responsible for checking the full outcome.
“protects the integrity of the technological system, at the expense of the nearest human operator.”
— Madeleine Clare Elish, as quoted in Meyer’s book
Questions the Book Cannot Settle
The source material does not establish how courts would assign responsibility in a specific dispute involving an AI agent. Legal outcomes depend on the facts, contracts and applicable law; Meyer’s discussion of liability limits and regulatory coverage is his account of the issues, not a ruling on any particular agent transaction.
The steel purchase is explicitly presented as a composite scenario, so it should not be read as an actual company’s loss. The supplied material also ends partway through its discussion of a 2023 Amsterdam Court of Appeal case, leaving the case’s full facts and the book’s treatment of it unavailable here. It does not provide the full text or jurisdictional scope of the vendor contract terms Meyer cites.
It is also unclear from the supplied information how widely the ebook has been distributed, how it has been received, or whether organizations or regulators will adopt the safeguards Meyer proposes. Publication alone does not change legal duties or determine who would be responsible after a real-world incident.
How Organizations May Respond
The ebook is listed as available through several retailers, including Apple Books, Barnes & Noble, Kobo, Thalia, Bookshop.org, Smashwords, Fable and Vivlio. Meyer also points readers to a short video introducing the book’s argument; the supplied description says its narration uses an AI voice clone of the author.
For organizations using agents, the immediate questions raised by the book are operational: which person owns each agent, what actions it may take, and how people can interrupt or challenge its decisions. Teams can also examine whether their logs record the relevant approvals and information—not just the final action—and whether an employee has a genuine opportunity to refuse a decision.
No policy change, court decision or regulatory action is announced in the source material. The next developments to watch are any further details from Meyer about the legal examples in the book, and whether organizations or lawmakers respond to the accountability gaps he describes.
Key Questions
What is Many Hands about?
Thorsten Meyer’s October 2026 ebook examines who pays, who explains and fixes an outcome, and who bears moral responsibility when an AI agent acts for a person or organization.
Is the steel purchasing story a real incident?
No. Meyer describes Theresa and the purchasing decision as a composite scenario built from common situations, not as a documented transaction at a named company.
Does the book establish who is legally liable for an agent’s actions?
No. It frames liability as a legal question shaped by the facts, contracts and applicable law. The supplied material does not report a new court ruling or settle liability for AI agents generally.
What safeguards does Meyer propose?
He recommends a named person responsible for each agent, permissions matched to its task, decision records and a way for affected people to raise concerns. He also argues that human approval requires adequate time, information, competence and authority to refuse.
Source: ThorstenMeyerAI.com
