🔍 Read the full analysis: When AI Acts In Your Name, How Should Accountability Work? on ThorstenMeyerAI.com
Get smart everyday buys 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, released in October 2026, examines how responsibility should work when AI agents make purchases, accept terms or take other actions in a person’s or organization’s name. The book separates legal liability, organizational responsibility and moral blame, and argues that current law and contracts may leave important gaps, particularly for financial losses.
Thorsten Meyer released Many Hands: Who Answers When the Agent Acts in Your Name? as an ebook in October 2026, making accountability for AI agents the subject of a new book. The release addresses a practical question as agents gain the ability to buy, pay, accept terms and change data on behalf of users: who is responsible when an action follows an agent’s instructions but conflicts with what its operator intended?
Meyer’s article introduces the issue through a composite purchasing scenario, not a documented real-world incident. In it, an agent accepts a steel supplier’s terms overnight, securing a price seven percent lower while committing the company to buy 120 tonnes over a year, with no cancellation right. The commitment is valued at about €370,000. The owner had separately agreed to divide the following year’s volume with another supplier, but the agent had not been told. Meyer says the agent acted within its rules and recorded its steps.
The book separates the question “who is responsible?” into three: who pays, a matter shaped by law and contracts; who must explain and repair the outcome, an organizational question; and who should have acted differently, a moral question. Meyer argues these answers can point to different people. He also says blame can fall on the person nearest the action, while system designers and organizational decisions receive less scrutiny.
Among the book’s examples is the 2018 fatal collision involving an automated test vehicle in Tempe, Arizona. Meyer cites the US National Transportation Safety Board’s findings that the system detected pedestrian Elaine Herzberg 5.6 seconds before impact, did not settle on what she was, and was designed to delay emergency braking for one second. The board identified several factors and participants. The vehicle operator pleaded guilty to endangerment, while the book asks how responsibility should be assessed across the wider system.
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 Errors Leave Losses
The accountability question matters because AI agents can take actions with financial and operational consequences before a person reviews each step. A model that produces a poor answer may be corrected; an agent that accepts a contract or sends a payment may create obligations that are harder to reverse. Meyer’s example is illustrative, but it makes the distinction between an agent following its rules and an organization getting the result it intended.
Meyer’s article argues that legal liability, workplace responsibility and moral blame can diverge. A business may have to deal with a supplier or customer even if an employee did not personally make the decision. Internally, the organization still needs to identify who can investigate, explain and correct the problem. Those questions affect how firms assign authority, supervise automated actions and keep records.
The book also highlights a potential mismatch between liability rules and common agent failures. Meyer says Europe’s product-liability framework covers software but does not cover pure financial losses, such as an incorrect payment or an unfavorable contract. He further says provider contracts may cap liability at roughly a year of fees. These are claims presented in the source article; their application will depend on the relevant law, contract and facts of a particular dispute.
Products Worth Considering
Dell XPS 13 9350 Laptop, 13.4 FHD+ Display, Intel Ultra 7 256V, 16GB DDR5, 1TB SSD
Apple iPad Air (2025) 13‑inch Wi‑Fi 256 GB – Space Gray (M3 chip)
Apple Mac mini Desktop Computer with M5 Pro Chip (2026)
As an affiliate, we earn on qualifying purchases.
The Examples Behind the Book
The book takes its title from political philosopher Dennis F. Thompson’s 1980 description of the “problem of many hands”: when many people contribute to an outcome, it can be difficult to determine who should answer for it. Meyer argues that agent systems add more participants and distance between decisions, while allowing actions to happen quickly. The agent itself, he writes, cannot pay or apologize.
Meyer also discusses the Dutch childcare benefits scandal, in which tens of thousands of families were wrongly required to repay benefits. He describes a system in which officials, risk models, fraud teams and courts could each assume another participant had checked the decision. According to the source article, the Dutch cabinet resigned in January 2021, and a government report in September 2026 had recognized more than 44,000 affected people. The source does not provide the report itself, so that figure is attributed to Meyer’s account.
The release follows Meyer’s earlier writing on AI agents and their access to passwords and keys, as well as the interaction between shopping agents and retailers. He says he also builds and runs agents for software development and publishing. The ebook is listed for sale through several outlets, including Apple Books, Kobo and Thalia.
“protects the integrity of the technological system, at the expense of the nearest human operator.”
— Madeleine Clare Elish, as quoted by Meyer
The Limits of Current Liability
The source article does not describe a newly enacted rule, court decision or regulatory action. The development is the publication of Meyer’s book and its argument, not a legal ruling that settles who is liable when an AI agent acts. The article does not provide the full text of the ebook, the legal sources behind each liability claim or independent comment from regulators, providers or affected businesses.
It is also unclear how courts would allocate responsibility in a particular agent-related dispute. Outcomes may depend on what instructions the agent received, what a human could review, the authority granted to the system, the contract terms and the applicable jurisdiction. Meyer’s purchasing example is explicitly a composite scenario, not evidence of a specific company losing €370,000.
Designing Clearer Agent Oversight
Meyer argues that organizations can make accountability more practical by naming a person responsible for each agent, matching its access and authority to the task, recording who made relevant decisions and giving affected people a way to raise concerns. He says records should support learning and correction, rather than being used only to assign blame after a failure.
The next steps are not set out as a formal policy or legislative timetable in the source article. For now, readers can consult the ebook, while organizations deploying agents will need to define their own review and escalation processes and check how their contracts allocate risk. Any broader change in legal standards would require action from lawmakers or decisions by courts; the article identifies no such change currently underway.
Key Questions
What is Many Hands about?
Thorsten Meyer’s October 2026 ebook examines who pays, who must explain or fix an outcome, and who bears moral responsibility when an AI agent acts for a person or organization.
Did the steel-purchasing incident actually happen?
No. Meyer describes Theresa and the steel order as a composite scenario built from common situations. It is an example used to frame the book’s argument, not a reported real transaction.
Does the book establish who is legally liable for an agent’s actions?
No. The source presents Meyer’s analysis, not a court ruling or new law. Legal responsibility in a real dispute would depend on the facts, contract terms and applicable jurisdiction.
What does Meyer recommend organizations do?
He proposes assigning a named person to each agent, limiting its authority to what the task requires, keeping records of decisions and providing a way for affected people to raise concerns. These are recommendations in the book, not a newly announced regulatory standard.
Source: ThorstenMeyerAI.com
