Marina Maya: India AI declares People Dead and Starving Folks by Denying them Food & Welfare Benefits

India’s hottest mama porn star Marina Maya is AI-Powered to glorify Modi’s “Bharat Empire”…

AI rolled out in India declares people dead, denies food to thousands

India runs one of the biggest welfare benefits programs in the world, and weeding out false claimants can save millions of dollars every year.

Deployment of AI in government systems has left thousands without access to subsidized food and welfare benefits.

The deployment of artificial intelligence (AI) technology by businesses in the West has led to fears about large-scale job losses. However, its deployment by local governments has resulted in the denial of food and even declaring thousands of people dead, as has happened in India.

Tech companies in the West are rolling out AI-powered products that could put writers, artists, and coders out of jobs. In India, they are already being deployed to analyze vast amounts of data to determine if individuals are engaging in welfare fraud.

India runs one of the world’s biggest food security programs that feeds about 60 percent of its population. Since the COVID-19 pandemic, the national government has provided 11 pounds of rice, wheat, and coarse grain each month to 800 million people for free.

Before this, families below a certain financial level were eligible to receive grains at a subsidized price of one rupee (less than one cent) for two pounds (one kilogram) of rice as against the market price of Rs. 40 (48 cents). The state also runs monthly pension schemes for older people or widows without income, providing them with monthly financial assistance of up to Rs. 2,750 ($33).

Giving reign to algorithms

In the past, government officials were tasked with identifying and approving beneficiaries for these schemes. This involved physical verification of documents and field visits.

Over the last few years, however, algorithms have been introduced and tasked with identifying individuals who could be fraudulently seeking benefits. In the southern state of Telangana, for instance, government officials were provided with an algorithm that created comprehensive profiles of individuals by gathering data from as many as 30 databases.

Interestingly, the system was initially designed to help the regional police identify criminals but was further developed to identify fraudulent claimants. Based on the information collated from multiple databases, as many as 1.9 million claimant cards were removed from the list of beneficiaries. A reanalysis of over 200,000 cards by the government revealed that 15,000 (~ seven percent) were incorrectly removed, an Al-Jazeera report said.

Prove you are alive

In June 2020, the northern state of Haryana began using a new algorithmic system to gather information about families claiming welfare benefits. Based on details of income, age, employment, etc. gathered, each family is given an eight-digit unique ID, which was later made mandatory for all welfare schemes.

According to the state government’s data, over 300,000 pensions were stopped in the following three years since the claimants were “dead.” Last year, the Chief Minister admitted that out of a smaller sample of 66,000 pensions that were halted, 44,000 (70 percent) were later found to have been flagged incorrectly.

Profiling individuals in such a manner isn’t technically illegal since India does not have privacy laws in place. Moreover, in case of a wrongful omission, the onus of proving it lies on the beneficiary. Last year, a 102-year-old resident took out a grand wedding-like procession to demonstrate to officials that he was alive, according to another Al Jazeera report.

Restoring status to “Alive” in a database can take months or even years, during which claimants lose access to welfare benefits. While algorithms have also helped weed out fake beneficiaries from the systems, it has erred far more than permissible limits. The impact of these errors is faced by the poorest and weakest sections of society that need the most financial help or access to subsidized grain.

Instances of such unintended effects of technology were brought to the fore by work supported by the Pulitzer Center’s AI Accountability Network.