Accountability

Who answers when an AI agent causes harm.

A Miami jury found Tesla 33% at fault for a fatal Autopilot crash and awarded $243 million, and the judge upheld the verdict in February 2026. In the EU, software and AI fall under strict product liability from December 9, 2026.

Damages in AI harm cases: Benavides v. Tesla $243M; EEOC v. iTutorGroup $365,000; Ultraman case, Guangzhou $1,400; Moffatt v. Air Canada about $600 DAMAGES IN AI HARM CASES Benavides v. Tesla $243M EEOC v. iTutorGroup $365,000 Ultraman case, Guangzhou $1,400 Moffatt v. Air Canada about $600 US DOLLARS, LOG SCALE
Tesla punitive damages
$200M
EU product liability deadline
Dec 9, 2026
ISO generative AI exclusions
3 forms
Registry incidents
6

Evidence

Courts in Canada, the US and China rejected "the AI did it" as a defense

Documented cases: accountability
DateCaseCountryWhat happenedSource
Feb 2024Moffatt v. Air CanadaCanadaTribunal rejected the argument that the chatbot was responsible for itselfSource
Aug 2025Benavides v. TeslaUnited States$42.6 million compensatory plus $200 million punitiveSource
May 16, 2025Mobley v. WorkdayUnited StatesNationwide age-bias collective certified against an AI screening vendorSource
Feb 2024Ultraman caseChinaFirst ruling holding a generative AI provider liable for outputSource
Jan 2026Hangzhou AI hallucination caseChinaCourt held AI is not a legal subject and generative AI is a service with fault-based liabilitySource

Developers

Nine of 13 developers signed the Seoul commitments; one of five Chinese developers did

Frontier developers
DeveloperEU code of practiceSeoul commitments
OpenAI United StatesYesYes
Anthropic United StatesYesYes
Google DeepMind United StatesYesYes
Meta United StatesNoYes
xAI United StatesPartial safety chapter onlyYes
Microsoft United StatesYesYes
Amazon United StatesYesYes
Mistral FranceYesYes
DeepSeek ChinaNoNo
Alibaba (Qwen) ChinaNoNo
Zhipu AI (Z.ai) ChinaNoYes
Moonshot AI ChinaNoNo
ByteDance (Seed) ChinaNoNo

Tag colors: green yes, amber partial, red no, gray not found.

Registry

Six registry incidents bear on accountability

Registry incidents where this discipline was tested in practice. Each links to its full record.

Registry incidents linked to this page
IncidentDateDeveloperRegulator action
Uber Test Vehicle Kills Pedestrian in TempeMar 2018Uber ATGArizona suspended Uber's testing on March 26, 2018. Yavapai County declined to charge Uber in March 2019. NTSB adopted its findings on November 19, 2019. The backup operator pleaded guilty to endangerment in July 2023 and received three years of supervised probation.
Air Canada Chatbot Invents a Refund RuleNov 2022Not disclosedOn February 14, 2024, the BC Civil Resolution Tribunal found negligent misrepresentation and ordered Air Canada to pay CA$812.02.
Cruise Robotaxi Drags a PedestrianOct 2023Cruise (General Motors)California DMV suspended Cruise's driverless permits on October 24, 2023. NHTSA fined Cruise $1.5 million on September 30, 2024, for incomplete crash reports. On November 14, 2024, Cruise admitted filing a false report and agreed to a $500,000 deferred prosecution agreement.
Packing Robot Kills a Worker in GoseongNov 2023Robot installer (not named)Goseong police investigated, including questioning the on-site safety manager. No ministry action was found.
Tesla Recalls Autopilot After NHTSA InvestigationDec 2023TeslaNHTSA opened PE21-020 in August 2021, upgraded it to EA22-002 in June 2022 and closed it in April 2024 after reviewing 956 crashes, opening RQ24-009 on whether the fix worked. No penalty.
Xiaomi SU7 Crashes on Assisted DrivingMar 2025XiaomiIn April 2025 MIIT barred carmakers from calling driver assistance autonomous driving and required approval of driving-function updates. On September 19, 2025, Xiaomi recalled 116,887 SU7s through SAMR over a highway driver assistance flaw.

Rules

The EU chose strict product liability; insurers are writing AI out of policies

Rules and frameworks: accountability
RuleJurisdictionWhat it requiresStatus
Product Liability Directive 2024/2853EUStrict liability for software and AI; defect presumptionTransposition by Dec 9, 2026
AI Liability DirectiveEUFault-based AI claimsWithdrawn Feb 2025
Interim Measures for Generative AI ServicesChinaProvider duty of care, complaints, labelingSince Aug 2023
ISO generative AI liability exclusionsUS insuranceExclude generative AI injury from general liabilityJan 2026 edition

Status colors: green in force, blue voluntary standard or framework, amber adopted and phasing in, gray draft or pending, red withdrawn.

Record

Page Record

Datasets
ai_trust_evidence.csv (5 rows); registry incidents (6 linked); developer scorecard (13 developers, static table)
Key sources
Moffatt v. Air Canada
Benavides v. Tesla
Mobley v. Workday
Ultraman case
Hangzhou AI hallucination case
Last verified
October 2026