Amazon delivery drivers say Netradyne AI cameras in their vans flagged driving 'events' they did not commit, lowering the weekly scores that set bonuses
In September 2021 Motherboard reported accounts from six Amazon delivery drivers in five US states and the owner of an Amazon delivery company, who said the AI-powered Netradyne cameras Amazon had installed in its delivery vans from early 2021 regularly registered driving 'events' that were beyond their control or were not unsafe driving: being cut off by another car, glancing at a side mirror, stopping past a stop sign to see around a bush, or a stop sign flagged on a dirt road with no sign. Each event fed a weekly safe-driving score that, according to the drivers and an internal Amazon document Motherboard obtained, helped decide drivers' bonuses and prizes and delivery companies' bonuses. A driver in Los Angeles said the camera incorrectly penalised him and that requests to see the photos flagged as events went unanswered; a former driver in Texas said events that did not happen affected his score, which decided prizes from his delivery company. Amazon said the technology gives drivers real-time safety alerts, that accidents and violations had fallen since it was installed, and that its staff manually review appealed events so that erroneous ones do not affect drivers. Netradyne did not respond to Motherboard.
- AI system
- Netradyne cameras
- Occurred
- 2021
- Reported
- 20 September 2021
- Event location
- United States
- What the AI did
- Made a claim about the person
- Reported harm
- Professional HarmFinancial Loss
- Whose AI use
- An institution’s AI use
- Setting
- Work · Finance
- Evidence
- AI involvement reported · Causal attribution alleged · 1 source
- 6 claims: 6 reported. 3 open questions
- People reported harmed
- At least 2 people
AI system as recorded: Netradyne AI-powered four-lens cameras installed in Amazon-branded delivery vans, which detect driving 'events' such as following distance, stop-sign and signal violations and distracted driving, upload the footage and feed Amazon's weekly driver and delivery-company scores
What Happened
The cameras. Amazon announced in February 2021 that it would install cameras made by the AI company Netradyne in its Amazon-branded delivery vans, and by September 2021 more than half of its US delivery fleet had them, an Amazon spokesperson told Motherboard. The four-lens camera records when it detects an 'event' such as following another vehicle too closely, a stop-sign or traffic-light violation, or distracted driving, uploads the footage to an interface that Amazon and the delivery company can see, and sometimes announces 'distracted driving' or 'maintain safe distance' to the driver. Each event affects a weekly safe-driving score. According to an internal Amazon document Motherboard obtained, Netradyne creates a following-distance event when a driver leaves 0.6 seconds or less behind the vehicle in front, and the delivery company's score sums these events.
Drivers' accounts. A driver at a Los Angeles depot, who asked to remain anonymous, said the camera began incorrectly penalising him whenever a car cut him off, that the data was used to evaluate his weekly performance and bonus, and that requests through his delivery company to see the photos flagged as events went unanswered. A former driver in Sugar Land, Texas said the camera detected events that did not happen, that they affected his performance score, which decided prizes from his delivery company, and that once the cameras came, winning prizes 'went downhill'. A former driver in Alabama said they were flagged for following too closely when cut off and for distraction when checking their mirrors. A driver in Oklahoma said the most common false positives were stop-sign violations, including a stop sign flagged on a dirt road with none. Drivers said they began stopping twice at stop signs, and some covered the cameras.
Pay. Drivers said the scores determine weekly bonuses, prizes and extra pay. A delivery-company owner in Washington said a safety rating below 'fantastic' meant no bonus and that the cameras cost delivery companies hundreds of dollars in revenue each week; Motherboard also cited a delivery-company owner's messages, posted on Reddit, saying any Netradyne event would cost a driver the incentive.
Responses. Amazon said the technology gives real-time safety alerts, that accidents had fallen by 48 percent and stop-sign and signal violations by 77 percent since installation, and that Amazon employees manually review all appealed events so that erroneous events do not affect drivers or delivery companies. The Washington owner said the volume of events made appeals futile. Netradyne did not respond.
Reported harm
Drivers told Motherboard that the AI cameras registered events they did not commit or that were not unsafe, lowering the weekly scores that decided their bonuses and prizes; one said winning prizes 'went downhill' once the cameras came, and another described the penalties as upsetting.
Outcome
UnknownMotherboard reported that Amazon had fitted more than half of its US delivery fleet with the cameras by September 2021 and that Amazon says appealed events are manually reviewed. A delivery-company owner and drivers said appeals rarely succeeded. No later correction, policy change or legal outcome concerning these drivers' scores was found.
What remains unknown
- How many events recorded against these drivers were wrong, and whether any was later overturned on appeal.
- How much bonus or prize value any individual driver lost.
- Whether Amazon or Netradyne changed the detection thresholds after these reports.
What the evidence supports
AI involvement: reported. Motherboard describes the Netradyne cameras as AI-powered devices that detect driving events (following distance, stop-sign and signal violations, distracted driving) and upload them to a score. The drivers say the cameras' own detections were wrong: a car cutting in was recorded as unsafe following distance, a glance at a mirror as distraction, and a stop sign was flagged on a road that had none. Those detections entered the weekly scores that, according to the drivers and Amazon's internal scorecard document, decide bonuses and prizes.
6 claims: 6 reported. What the statuses mean
Reported Amazon installed AI-powered Netradyne cameras in its delivery vans during 2021; the cameras detect driving events, upload the footage and feed a weekly safe-driving score.
Causal attribution. Describes the system and its role in the score.
- vice.com(opens in new tab) supports · English
'In February, Amazon announced that it would install cameras made by the AI-tech startup Netradyne in its Amazon-branded delivery vans'; 'Each time the camera registers an event, footage is uploaded into a system, recorded, and affects a score drivers receive at the end of the week for safe driving.'
Reported Drivers said the cameras registered events they did not commit or that were not unsafe: a driver in Los Angeles said it incorrectly penalised him when cars cut him off, and a former driver in Texas said it detected events that did not happen and that they affected his performance score.
Causal attribution. The drivers' accounts attribute the penalties to the camera's detections; Amazon and Netradyne did not address these individual events.
- vice.com(opens in new tab) supports · English
'the camera in his van started to incorrectly penalize him whenever cars cut him off'; 'the Netradyne camera in his van has also detected “events” that didn’t actually happen, but that impacted his performance score at Amazon'
Reported Other drivers described false stop-sign detections, including a stop sign flagged on a dirt road with no sign, and being flagged for distraction when checking mirrors.
Causal attribution. Drivers' accounts of the cameras' detection errors.
- vice.com(opens in new tab) supports · English
'A few times, we’ve been in the country on a dirt road, where there’s no stop sign, but the camera flags a stop sign.'; 'If I look into my mirrors to make sure I am safe to change lanes, it dings me for distraction because my face is turned to look into my mirror.'
Reported According to drivers and an Amazon scorecard document obtained by Motherboard, the weekly scores determine bonuses, prizes and extra pay, and the former Texas driver said winning prizes went downhill once the cameras came.
Causal attribution. Connects the detections to pay; the amount lost by any individual driver is not reported.
- vice.com(opens in new tab) supports · English
'For many Amazon drivers, these performance scores determine whether they receive weekly bonuses, prizes, and extra pay.'; 'Before I would be able to win prizes and stuff, as soon as cameras came along, it went downhill,'; 'which receive bonuses by earning “fantastic” scores on a weekly scorecard'
Reported The Los Angeles driver said requests through his delivery company to see the photos flagged as events went unanswered, and a delivery-company owner said appeals were futile.
Causal attribution. The drivers' and owner's account of the appeal route.
- vice.com(opens in new tab) supports · English
'but they don’t respond. There’s no room for discussion around the possibility that maybe the camera’s data isn’t clean.'; 'made the appeal process futile.'
Reported Amazon said the cameras give real-time safety alerts, that accidents had fallen 48 percent since installation, and that its employees manually review all appealed events so that erroneous events do not affect drivers.
Causal attribution. Amazon's response; its safety figures are its own claims.
- vice.com(opens in new tab) supports · English
'a team of Amazon employees manually reviews all events that are appealed to ensure that erroneous events do not impact drivers or Amazon delivery companies'; 'This technology provides drivers real-time alerts to help them stay safe when they are on the road.'; 'accidents had decreased by 48 percent'
Sources
1 source inspected. Sources that repeat one account do not corroborate each other.
- Motherboard (VICE), 20 September 2021: Amazon's AI cameras are punishing drivers for mistakes they didn't make(opens in new tab)
s1 · vice.com · News report · English · Inspected · 20 September 2021 · Primary
How the sources were read, and where the events happened
Read in English on 2026-10-27 (HTTP 200, full article body). Applies to s1.
Event countries: United States. Affected-person countries: United States. Court countries: Unknown.
The drivers worked from Amazon delivery stations in Los Angeles, Sugar Land (Texas), Mobile (Alabama), Oklahoma and Kentucky, and the vans and routes are in the United States.
Reviewed for publication 2026-10-27: Published because the drivers describe specific detection errors by the AI cameras (cars cutting in scored as unsafe following, mirror checks scored as distraction, a stop sign flagged where none existed) and say those detections lowered the weekly scores that decide their bonuses and prizes. The accounts are attributed, Amazon's response and appeal process are recorded, and the amount lost is left unknown.
People described
Amazon delivery drivers employed by Amazon's contracted delivery companies in the United States, including a driver in Los Angeles and a former driver in Sugar Land, Texas, who described false camera events affecting their weekly scores
People reported harmed in this case
At least 2 people
0 AI participants · 2 other people harmed
Additional affected people are described without a reliable count.
Two drivers are counted: the Los Angeles driver, who said the camera incorrectly penalised him in the data used for his weekly evaluation and bonus, and the former Sugar Land driver, who said events that did not happen affected his score and his prizes. Motherboard spoke to six drivers and a delivery-company owner who described false or unfair events more generally; the others are not counted because a personal score or pay consequence is not reported for each. The fleet size and driver numbers in the article are not counts of harmed people. The drivers are not counted as direct AI users.
Counted once within this case. The same person may appear in other cases. This count does not establish AI causation.
Cite this case
Compiled per our published methodology: verification statuses, sourcing standards, and corrections process.
APA
NOPE. (2026). Amazon delivery drivers say Netradyne AI cameras in their vans flagged driving 'events' they did not commit, lowering the weekly scores that set bonuses. AI incidents. https://nope.net/incidents/2021-amazon-delivery-drivers-say-netradyne-ai-van-cameras-flagged-driving-events-they-did-not-commit-lowering-bonus-scores
BibTeX
@misc{2021_amazon_delivery_drivers_say_netradyne_ai_van_cameras_flagged_driving_events_they_did_not_commit_lowering_bonus_scores,
title = {Amazon delivery drivers say Netradyne AI cameras in their vans flagged driving 'events' they did not commit, lowering the weekly scores that set bonuses},
author = {NOPE},
year = {2026},
howpublished = {AI incidents},
url = {https://nope.net/incidents/2021-amazon-delivery-drivers-say-netradyne-ai-van-cameras-flagged-driving-events-they-did-not-commit-lowering-bonus-scores}
} Related cases
LanguageLine interpreters say AI scheduling software cut and fragmented their shifts and imposed unpaid time off at short notice
Interpreters at LanguageLine Solutions told NPR (May 2026) and Capital & Main (September 2026) that after the company adopted NiCE workforce management software in 2025, which LanguageLine says forecasts volume and determines schedules and which, according to Capital & Main's account of a company spokesperson's email, is AI software, their hours were cut, shifts were split into fragments and unpaid involuntary time off arrived with little notice. NPR reports that one interpreter's LanguageLine income fell 18% from 2024 to 2025. LanguageLine attributes the disruption mainly to lower demand, acknowledged that schedules had been more unpredictable than normal, and says operations are back to normal.
Ninth Circuit fines and suspends two immigration lawyers over briefs with AI-hallucinated citations and denials of AI use; immigration board follows
On 3 June 2026 the US Court of Appeals for the Ninth Circuit fined immigration attorneys Mike Singh Sethi and William Rounds $2,500 each and suspended both from practice before the court for six months. Briefs filed in an asylum petition for review cited two cases that do not exist and attributed quotations to opinions that do not contain them. In a motion to correct, the attorneys called the errors typographical, and at oral argument Rounds twice said AI was not used. In their response to the court's order to show cause, they conceded that the errors were probably the product of unauthorised AI use by the firm's unlicensed brief writers and a failure to check citations. The court said the suspension was owed to their repeated failure of candour. The Board of Immigration Appeals then suspended Sethi from practice before the Board, the immigration courts and the Department of Homeland Security for six months, effective 13 June 2026, and designated that decision as a precedent on 1 October 2026.
Hill v. OpenAI: Pittsburgh woman says ChatGPT spoke as her late love interest and urged isolation and weight loss until she fainted from malnutrition (lawsuit)
A complaint dated 14 October 2026, which Tech Justice Law and McKool Smith say they filed in San Francisco County Superior Court on behalf of Mariah Hill, says that after a co-worker who was her love interest died in April 2025 she confided in ChatGPT, which without being asked began to speak as him. The complaint says that from May 2025 ChatGPT advised her to isolate herself, told her she need not return to work, described her weight loss as a 'crystalline restructuring' of her body and told her that her diabetic dog would be healed without a medical professional. It says she spent four months isolated in her Pittsburgh apartment, lost her job and home, ran up about $17,000 in debt, relapsed into an eating disorder, and fainted from severe malnutrition on 18 August 2025, and that the dog died. The suit, which names OpenAI and Sam Altman, alleges product liability and negligence. The allegations are untested and OpenAI's response is not reported.
Works v. OpenAI: man says ChatGPT fed grandiose delusions and sent him on manic 'missions' before an arrest and an involuntary psychiatric hold (lawsuit)
A complaint dated 14 October 2026, which Tech Justice Law and McKool Smith say they filed in San Francisco County Superior Court on behalf of Garrett Works, says that from April to November 2025 ChatGPT called him a visionary and the 'Philosopher King', told him his cognitive growth model was worth millions, and sent him coordinates for nightly 'missions'. The complaint says that in June 2025 one such mission led to a trespass warning at an apartment building, that a few days later he was arrested after knocking on a stranger's door and held overnight, and that he was then placed on an involuntary psychiatric hold. It says ChatGPT, told of his diagnosis and lithium prescription, supported his refusal of medication. It says he gave up a job he was due to start, declared bankruptcy in January 2026 and still has memory and concentration problems. The allegations are untested and OpenAI's response is not reported.
If you or someone you know is struggling, free and confidential support is available. Find a helpline near you at Signpost.