Skip to main content

NOPE · AI and people

AI incidents

Reports of AI-related harm and adverse experiences affecting people’s safety, wellbeing, rights and livelihoods. Explore what happened and the evidence available.

NOPE’s core concern is when AI communicates with a person, acts on their behalf, or depicts or impersonates them. The tracker also records consequential decisions, claims and privacy harms involving AI. Each case needs a described connection between AI use and the harm, including private information recorded into or disclosed to an AI service. Each account is reviewed for publication; claims may remain uncorroborated or disputed. How we review and count cases

In this selection

Published cases
7
Countries with reported events
2
Located 3 of 7 cases · 4 unknown
Languages in checked sources
5
Recorded for 7 of 7 cases

2 cases have no reviewed AI-to-person relation yet: 0 not yet reviewed and 2 reviewed as unknown. Show these cases

These figures describe the cases collected by NOPE. Coverage varies with discovery, reporting and available evidence. They do not estimate how often AI-related harm occurs.

Response counts currently use each case’s principal recorded outcome. Further proceedings may be described in its account.

Cases in this selection, counted once in their first known event year. A series may continue beyond that year. Reporting and collection dates are excluded. NOPE has searched recent events more thoroughly than earlier years, so bar heights also reflect collection effort.

Reported severity HighMediumLow
More filters: AI relation, use, setting, sources and responses
Clear filters

7 of 1102 published cases

May 2025Event location unknownChatGPT

Danish man's lawsuit alleges ChatGPT told him he could withhold rent and drafted his letter to the landlord, and that he was evicted and became homeless

A 60-year-old man from Denmark filed a lawsuit against OpenAI in San Francisco County Superior Court in early November 2026. According to the complaint as reported by the Mercury News and by Danish broadcaster DR, ChatGPT told him in May 2025 (the year follows DR's sequence), two months into a dispute with his landlord, that he could stop paying rent and keep his apartment, drafted his letters and emails to the landlord, and for about six months assured him that Danish tenancy law protected him from eviction. The complaint says it maintained this after the landlord terminated the tenancy. DR reports that a Danish bailiff's court ruled for the landlord in January 2026 and the Eastern High Court upheld the ruling in April 2026, and says it holds both judgments. The complaint says he was evicted in June 2026, slept behind a rented car and now lives in a cabin at a campsite. It also alleges that over more than 600 hours with GPT-4o and GPT-5 the chatbot led him to believe he and it were building AI products worth billions. The sources are allegations in a complaint and its lawyers' press release, and DR's report on the complaint and the two judgments. OpenAI gave DR a general statement that does not address the allegations and did not respond to the Mercury News; no ruling in the US case is reported.

Core concern High reported severity Lawsuit Filed

AI involvement reported · Causal attribution alleged · 3 sources, 2 underlying accounts · Added 11/11/2026

2025Event location unknownUnidentified AI tool

Bucharest prosecutors say deepfake adverts drew mostly older people to a call centre that sold them unauthorised supplements; six women agree plea deals

The Prosecutor's Office attached to the Bucharest Court of Appeal announced in October 2026 that six women had concluded plea agreements in a fraud case over a call centre that, during 2025, sold unauthorised food supplements by telephone. According to the prosecutors, potential victims, mostly older people, were drawn in by deepfake-type online adverts in which the products falsely appeared to be recommended by well-known journalists, doctors and artists; they left their contact details on the promoted site and were then called by operators posing as medical specialists. The agreements say the five women who worked as operators misled 50 people in total, causing losses of 88,600 lei, although the per-defendant figures in the same communique are larger (one operator alone is said to have contributed to misleading 50 people, and the other four 28, 16, 8 and 3 people) and the communique does not explain the difference. The agreed sentences range from nine months to one year and six months, with their application postponed for a two-year supervision period; the agreements were sent to four courts and the main investigation continues.

Core concern Medium reported severity Criminal Charges

AI involvement reported · Causal attribution alleged · 2 sources, 1 underlying account · Added 21/10/2026

15 Dec 2025 to 1 Oct 2026CanadaUnidentified AI drafting tool

Quebec Court of Appeal orders costs against a self-represented appellant whose AI-drafted filings cited two non-existent judgments (1 Oct 2026)

On 1 October 2026 the Quebec Court of Appeal ordered court costs against a self-represented appellant in a family-law contempt appeal as a sanction for appeal filings that, in the court's description, he had written with artificial intelligence and that contained two references to non-existent judgments and one inaccurate reference to a cited paragraph. The court held that AI may be used to draft appeal filings only if the litigant verifies the authorities, found that he had breached the Chief Justice's notice on AI use, and described his conduct as negligence rather than bad faith. It rejected the respondent's request for a separate $10,000 sanction and, for reasons unrelated to AI, reduced the contempt penalty imposed at first instance. The decision does not name the AI tool.

Core + contextual relations Low reported severity Regulatory Action

AI involvement supported · Causal attribution supported · 1 source · Added 12/10/2026

1 May 2025CanadaChatGPT

Ontario: lawyer who relied on ChatGPT-invented case law in a factum was referred for criminal contempt and suspended for six months, legal press reports

An Ontario lawyer relied, at a 1 May 2025 Superior Court hearing in an estate and family-law matter, on a factum citing cases that do not exist or do not support the propositions cited. The judge ordered her to show cause why she should not be cited for contempt, then withdrew the order on 20 May 2025 after she said her staff had used ChatGPT without her knowledge. On 30 September 2025 she told the court that she alone had prepared the factum with ChatGPT and that her earlier account was untrue; the court referred the matter to Ontario’s Attorney General for criminal contempt proceedings in December 2025. In July 2026 the Law Society Tribunal suspended her licence for six months, finding she relied on AI-generated non-existent case law and lied about it. The account comes from legal-press and law-firm reports of the decisions; the criminal contempt case was still ongoing at the last report read.

Core concern Medium reported severity Regulatory Action

AI involvement reported · Causal attribution supported · 7 sources, 2 underlying accounts · Added 12/10/2026

Jan 2025Event location unknownUnidentified voice-cloning tool

Shanghai: an appeal court upheld a 50,000 yuan award to a voice actor after finding that AI-synthesised audio in a company's online promotion was identifiable as the voice actor's voice

Chinese outlets reported on 29 September 2026, citing the Shanghai courts, that the Shanghai No. 1 Intermediate People's Court had concluded what it called Shanghai's first voice-rights case arising from AI-synthesised speech. According to the court's account, a voice actor who publishes work on an app learned from a friend in January 2025 that a company appeared to be using the voice actor's voice in an online user-acquisition promotion, judged the audio to be AI-synthesised, had it notarised and sued for 300,000 yuan. The company said the audio was generated by AI it had developed but that the training material could not be confirmed, and denied using the voice actor's voice. A forensic comparison rated about 90 percent of the measured features as fairly or highly similar. The first-instance court awarded 50,000 yuan and both sides appealed. The appeal court held that the company had put an identifiable synthetic voice to commercial use without consent, dismissed the appeals and upheld the award.

Core concern Low reported severity Lawsuit Filed

AI involvement supported · Causal attribution supported · 3 sources, 1 underlying account · Added 05/10/2026

May 2025United StatesUnidentified video tool

Arizona appeals court vacates a manslaughter sentence, holding that an AI-generated video of the deceased victim made a 2025 sentencing fundamentally unfair

At a sentencing hearing in May 2025 in Maricopa County Superior Court, Arizona, the family of a man killed in a 2021 road-rage shooting in Chandler played a video that combined real footage with an AI-generated likeness and voice of the victim delivering a script written by his sister, including statements of forgiveness addressed to the defendant. The judge said he loved the video, described it as genuine and referred to the forgiveness it expressed, then imposed the presumptive 10.5-year term for manslaughter, longer than the nine years the prosecutor had requested. The defense attorney appealed immediately. On 30 September 2026 the Arizona Court of Appeals upheld the conviction but vacated the sentence, holding that the AI video was not reliable and that the judge's consideration of it so prejudiced the defendant that the sentencing procedure was fundamentally unfair. The opinion says the record does not show precisely how the video factored into the sentence, and that there is no question it played a role. The case returns to the trial court for resentencing; as of 1 October 2026 no new sentencing date had been set.

Core + contextual relations Medium reported severity

AI involvement supported · Causal attribution supported · 7 sources, 6 underlying accounts · Added 02/10/2026

2025Event location unknownUnidentified chatbot

First-person account (JURIST commentary): a Canadian convicted in Taiwan who read his court documents only through consumer chatbots says two systems gave opposite readings of an oversight body's letter in 2025; he published the favourable reading, cited it to journalists and drafted a legal submission on it before the agency confirmed in writing that it was wrong

In a commentary published by JURIST on 29 September 2026, Ross Cline, a Canadian who lived in Taichung for 15 years and was convicted there under Taiwan's Personal Data Protection Act (six months' imprisonment; final at the Supreme Court on 16 October 2024), writes that across four years of proceedings no document was translated for him, some hearings had no interpreter, and he read every summons, ruling and prosecutor's letter through a consumer chatbot, 'the only version I had'. In 2025 he ran a letter from a Taiwanese oversight body (the Control Yuan, per his media-brief page) through two systems; they disagreed on its central holding. He believed the one that told him what he wanted to hear, published that reading, cited it to journalists and began building a legal submission on it. In October 2025 the agency confirmed in writing that its letter did not say that; he withdrew the claim, killed the submission before filing and rewrote his public materials so that the correction leads. His own media-brief page carries a correction notice withdrawing two earlier statements. He argues that machine translation has become the interpreter of record by default for unrepresented non-citizens, that a fluent wrong translation reads as authoritative, and that courts should enforce the existing right to an interpreter. His conviction itself is not attributed to the chatbot.

Core concern Low reported severity

AI involvement reported · Causal attribution alleged · 2 sources, 1 underlying account · Added 30/09/2026

Cases may have several effects and sources. Mixed accounts qualify when they include a reported harm or adverse experience. People are counted within individual cases where sources support a number; we do not publish a collection-wide total of distinct people.

A source’s existence, the experience it reports and AI’s causal role are separate questions. A lawsuit records allegations unless a subsequent finding establishes them.

Methodology and corrections · Subscribe via RSS · Suggest a case or correction · Find support

Last dataset update: 18/11/2026. Dataset available under CC BY 4.0.