Skip to main content
NOPE
Medium reported severity Media Coverage

Al Jazeera investigation says Telangana's Samagra Vedika system wrongly matched poor families to car owners, and officials denied them food subsidy cards

An Al Jazeera investigation published on 24 January 2024, produced with The Reporters' Collective and the Pulitzer Center's AI Accountability Network, reports that Samagra Vedika, a system the Telangana government uses to consolidate residents' data and check welfare eligibility, wrongly tagged poor families in Hyderabad as owning cars, and that officials relied on those tags to deny them food security cards (subsidised rice). Al Jazeera says the system uses machine learning and entity resolution to match people across databases, and that since 2016 officials have been mandated to check whether the system approved an applicant before deciding. In one case it reports that the system mistook a widow's late husband, a rickshaw puller, for a car owner with a similar name; her renewal application was rejected in 2021 for possessing a four-wheeler. After she traced the vehicle's real owner, officials agreed there had been an algorithmic error but said her family income exceeded the limit, which Al Jazeera says was not the case; she had been without the subsidised rations for more than seven years. A second family, whose husband cannot drive, was rejected in 2021 for possessing a four-wheeler; in November 2023 the High Court of Telangana said the petitioner is eligible whenever new cards are issued. The government says the technology is highly precise and that wrong vehicle or house tags occur in under five per cent of cases and have mostly been corrected.

AI system
Samagra Vedika
Occurred
2021
Reported
24 January 2024
Event location
India
What the AI did
Made a claim about the person · Informed a human decision about the person
Reported harm
Financial LossOther Material Harm
Whose AI use
An institution’s AI use
Setting
Public services · Finance
Evidence
AI involvement supported · Causal attribution supported · 1 source
7 claims: 1 disputed, 6 reported. 4 open questions
People reported harmed
At least 2 people

AI system as recorded: Samagra Vedika, the Telangana government's data-integration and profiling system developed by Posidex Technologies, which Al Jazeera says uses machine learning and entity resolution technology to identify people across government databases and tag welfare applicants as ineligible

Reported harm

According to an Al Jazeera investigation, Telangana's Samagra Vedika system wrongly tagged poor families as car owners and officials denied them food security cards on that basis; a widow had been without subsidised rations for over seven years and another family pays market prices for rice.

Outcome

Ongoing

Al Jazeera reports that the High Court of Telangana said in November 2023 that the second family's petitioner is eligible whenever new cards are issued, and that the widow was still waiting in January 2024. In April 2022 the Supreme Court ordered field re-verification of cards deleted since 2016 in a petition filed for excluded families. No later outcome was found on 2026-10-24.

What remains unknown

  • Which three petition cases Al Jazeera verified, and how many wrongful exclusions the system itself caused.
  • Whether the widow's card was later issued, and whether the second family has received a card since November 2023.
  • How the matching works in detail; the state withheld the source code and data formats.
  • Why the widow's card was first cancelled in 2016.

What the evidence supports

AI involvement: supported. Al Jazeera reports that Samagra Vedika matched a widow's late husband to a car owner with a similar name and tagged her family as owning a four-wheeler, that officials (mandated to check the system's approval before deciding) accepted the tag and rejected her food security card in 2021, and that they later agreed there had been an algorithmic error. It describes the system as using machine learning and entity resolution to match people across databases; the government describes it as using big data and disputes the error rate. The source code and data formats were withheld from the reporters.

7 claims: 1 disputed, 6 reported. What the statuses mean

Reported Since 2016 Telangana has used Samagra Vedika, a system that consolidates residents' data from several government databases, to check welfare eligibility; officials processing food security cards were mandated to check whether the system approved an applicant before deciding, and could depart from it only by giving reasons and evidence.

Causal attribution. Background on the system and how its output entered officials' decisions.

  • aljazeera.com(opens in new tab) supports · English
    'started arming the officials with Samagra Vedika, an algorithmic system that consolidates citizens'; 'the officials were mandated to check whether the algorithmic system approved the eligibility of the applicant before making their own decision'; 'They could either go with the algorithm'
Reported Al Jazeera reports that Samagra Vedika uses machine learning and entity resolution technology to identify a person across databases, and tags a claimant as ineligible when it finds a close enough match violating a criterion such as owning a four-wheeler.

Causal attribution. The description of the technology is Al Jazeera's, quoting terms whose origin the article does not state; the government describes it as using big data.

  • aljazeera.com(opens in new tab) supports · English
    'Samagra Vedika uses'; 'to uniquely identify a person'; 'If it finds a match close enough violating any of the criteria, it tags the claimant as ineligible'
Reported According to Al Jazeera, the system mistook a widow's late husband, a rickshaw puller, for a car owner with a similar name; her food security card renewal application was rejected in 2021 because she was said to possess a four-wheeler, and that the authorities believed their algorithm rather than her.

Causal attribution. Al Jazeera attributes the rejection to the system's wrong match and officials' acceptance of it.

  • aljazeera.com(opens in new tab) supports · English
    'mistook her late husband'; 'and the authorities accepted the algorithm'; 'finally rejecting it in 2021 because they claimed she possessed a four-wheeler'; 'The authorities did not trust her'; 'they believed their algorithm'
Reported After she traced the vehicle's real owner, officials agreed there had been an algorithmic error but said that her application could not be considered because family income exceeded the limit, which Al Jazeera says was not the case; she had been denied the rations for more than seven years when the article was published.

Causal attribution. The officials' acknowledgement of an algorithmic error is reported by Al Jazeera; the continued denial rests on a different, disputed ground.

  • aljazeera.com(opens in new tab) supports · English
    'found its real owner'; 'the officials agreed there was an algorithmic error but said that her application could not be considered because her total family income exceeded the eligibility limit even though that was not the case'; 'who has been denied ration for more than seven years, is still waiting'
Reported A second family, whose husband cannot drive, applied in 2018 and was rejected in 2021 for possessing a four-wheeler they do not own; Al Jazeera's caption says an algorithm tagged the family as owning a car, and the woman says the family pays more for the same rice bought from dealers. In November 2023 the High Court of Telangana said the petitioner is eligible whenever the state issues new cards.

Causal attribution. The algorithmic tag is stated in an Al Jazeera photo caption credited to The Reporters' Collective; the High Court remark is reported, not read in the order itself.

  • aljazeera.com(opens in new tab) supports · English
    'an algorithm has tagged the family as owning a car, denying them access to subsidised food grains'; 'The family applied for a food security card in 2018 but was rejected in 2021'; 'Instead of three rupees, we spend 10 for every kilogramme of rice'; 'the petitioner is eligible'
Reported Al Jazeera says it visited three of the roughly ten wrongful-exclusion cases listed in a Supreme Court petition and confirmed they were wrongly tagged by Samagra Vedika.

Causal attribution. Al Jazeera's own verification; the three cases are not identified.

  • aljazeera.com(opens in new tab) supports · English
    'The petition lists around 10 cases of wrongful exclusions'; 'Al Jazeera visited three of those and can confirm they were wrongly tagged by Samagra Vedika'
Disputed The Telangana government has said the technology is of high precision, that wrong tagging of a vehicle or house occurs in less than five per cent of cases and has mostly been corrected, and an official said verifications found very high levels of accuracy.

Causal attribution. Government counter-position; it does not address the individual cases.

  • aljazeera.com(opens in new tab) supports · English
    'The government has previously said that the technology is of'; 'It has also said that the error of tagging a vehicle or a house to the wrong person happens in less than five percent of all cases'; 'very high levels of accuracy'

Sources

1 source inspected. Sources that repeat one account do not corroborate each other.

How the sources were read, and where the events happened

Read in English on 2026-10-24 (full article). Reported by Tapasya, Kumar Sambhav and Divij Joshi with The Reporters' Collective and the Pulitzer Center's AI Accountability Network; the photo captions are courtesy of The Reporters' Collective. Applies to s1.

Event countries: India. Affected-person countries: India. Court countries: India.

Al Jazeera reports from Hyderabad that both families live there and applied to the Telangana government; the Supreme Court of India and the High Court of Telangana heard related petitions.

Reviewed for publication 2026-10-24: Published as a concrete adverse public decision informed by an automated matching system: Al Jazeera reports that Samagra Vedika wrongly matched a widow's late husband to a car owner and tagged another family as owning a car, and that officials rejected their food security cards on those tags; officials later acknowledged an algorithmic error in the widow's case.

People described

Low-income families in Hyderabad, Telangana, applying for or renewing food security cards, including a widow and a family whose husband cannot drive

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 applicants described with their rejections: the widow and the woman whose family was rejected in 2021 (members of their households are not counted separately). Al Jazeera says it visited three of about ten petition cases and confirmed they were wrongly tagged, without listing them, and gives state re-verification figures that it does not attribute to the system alone; these are not counted.

Counted once within this case. The same person may appear in other cases. This count does not establish AI causation.

Tags

#welfare-eligibility-algorithm#entity-resolution#food-security#wrong-match#public-services

Cite this case

Compiled per our published methodology: verification statuses, sourcing standards, and corrections process.

APA

NOPE. (2026). Al Jazeera investigation says Telangana's Samagra Vedika system wrongly matched poor families to car owners, and officials denied them food subsidy cards. AI incidents. https://nope.net/incidents/2021-telangana-samagra-vedika-algorithm-wrongly-tagged-poor-families-as-car-owners-food-security-cards-denied

BibTeX

@misc{2021_telangana_samagra_vedika_algorithm_wrongly_tagged_poor_families_as_car_owners_food_security_cards_denied,
  title = {Al Jazeera investigation says Telangana's Samagra Vedika system wrongly matched poor families to car owners, and officials denied them food subsidy cards},
  author = {NOPE},
  year = {2026},
  howpublished = {AI incidents},
  url = {https://nope.net/incidents/2021-telangana-samagra-vedika-algorithm-wrongly-tagged-poor-families-as-car-owners-food-security-cards-denied}
}

Related cases

Low Claude Code

Heritage project reports loss of inscription records after a Claude Code command

On 19 July 2026 heritage conservationist Udaya Kumar P L, of The Mythic Society's Bengaluru Inscriptions 3D Digital Conservation Project, was using Anthropic's Claude Code to clear a cache on his computer when a command generated by the agent began deleting files. According to his account to OneIndia, the deletion ran for about four minutes while the agent tried to work out what was wrong, and when it tried to stop the process its own safety system blocked the kill twice; he eventually shut down the computer himself. Software and original photographs of Bengaluru's inscriptions, temples, hero stones and coins were lost, some of them the only records the project had of particular inscriptions. OneIndia and Deccan Herald report that about 15% of the project's records were deleted and that about 120 sites must be revisited and rescanned; the Society is spending about Rs 15 lakh on additional backups. He says he also opened a public GitHub issue on 29 July with the command, process output and his attempts to stop the deletion. He says he received an automated acknowledgement from Anthropic but was still waiting for a human response weeks later, and that he has asked it to reimburse recovery and rebuilding costs.

Low Gemini (reported)

Ahmedabad, Gujarat: police say AI videos made from a businessman's photo passed Aadhaar facial authentication and a Rs 25,000 loan was taken in his name

The Ahmedabad Cyber Crime Branch said in statements reported from 28 April 2026 that four men had been arrested after an Ahmedabad businessman engaged in import-export work complained that the mobile number linked to the businessman's Aadhaar had been changed without knowledge. ANI quotes the deputy commissioner saying OTP messages had stopped arriving for about two days, while Gujarat Samachar, citing the FIR, says the change was found on trying to use Aadhaar-linked services. Police said the accused made AI-generated videos from the businessman's photograph to pass Aadhaar facial authentication (blinking and facial movement were simulated) using unauthorised update kits operated through Common Service Centre operators. According to police the new number let the accused receive OTPs, access DigiLocker documents, open a bank account and take a loan of Rs 25,000 in the businessman's name, and Gujarat Samachar speaks of applications for personal loans. Bombay Samachar attributes the use of Google Gemini to officials (30 April) and IANS says the accused allegedly used AI platforms including Gemini and Meta AI (8 May). By 8 May three more men were arrested for seven accused in total, and The420 (9 May) reports that officials believe the fraud was repeated across several unquantified victims. Every source rests on police statements and no source dates the event. The businessman and the accused are not named.

Low Cursor support bot (reported)

First-person post: Cursor subscriber says the company's AI support bot cancelled their prepaid annual plan when asked about a prorated refund

In a Reddit post on 22 October 2026, a subscriber to Cursor writes that they emailed Cursor's customer service asking whether a prorated refund was available for an annual subscription they had paid in full, and that the company's AI support bot, Sam, replied that it could not do that but could cancel the plan immediately. In a comment the poster quotes the bot saying it had cancelled the annual Pro subscription immediately, that Pro access had ended, and that the subscription did not meet refund eligibility requirements. The poster says the bot seems to have misunderstood the request, that the plan was downgraded to Free with more than six months left, that a request for a human produced a new ticket that never arrived, and that they would try a credit-card dispute. In April 2025 a Cursor co-founder wrote that the company used AI-assisted responses as the first filter for email support. No reply from Cursor staff appears in the thread. The account is uncorroborated.

Low Unidentified voice-cloning tool

Jewish Autonomous Region, Russia: police say scammers cloned a man's voice from old voice messages to ask his contacts for money; one paid 15,225 roubles

The Ministry of Internal Affairs directorate for Russia's Jewish Autonomous Region said on 21 October 2026 that residents increasingly report a combined scam in which messenger accounts are hacked, the chat history deleted, and the victim's contacts called with a voice cloned by neural networks. In the case it describes, a man in Birobidzhan followed a voting link sent from an acquaintance's hacked account. The scammers then gained access to his accounts and called his contacts; according to the police, they imitated his speech using voice-cloning technology based on his previously recorded voice messages. Seven of his acquaintances were misled and one transferred 15,225 roubles. The scammers also gained access to his account on the Gosuslugi state services portal. The police do not say how they established that a cloned voice was used, name a tool, or date the events.

If you or someone you know is struggling, free and confidential support is available. Find a helpline near you at Signpost.

Back to all incidents