Nov 21, 2026
Nov 21, 2026

The machine's fault, the citizenry's the one to suffer!
Governments today are claiming to leverage technology for faster, transparent, and effective governance. Artificial intelligence is being touted as a new tool for administrative reform, public service, and corruption control. From answering citizen queries in government offices to identifying beneficiaries of schemes, analyzing complaints, and making decision-making processes, AI is rapidly gaining ground. But amidst the dazzling hype surrounding technology, a fundamental question is being obscured: if a machine makes a wrong decision, which door will citizens turn to for justice? The use of AI in governance is certainly a matter of convenience, but it cannot be turned into a laboratory free of accountability.
The most dangerous tendency of generative AI is that it can confidently present false information. It can cite rules, facts, figures, or eligibility criteria that do not exist in reality. The average citizen may mistake the machine's answer for government authorization. If an elderly person is misinformed about pension eligibility, a student is misled about a scholarship, or a farmer receives incorrect information about a government scheme, the harm is not merely a technical error; it becomes an attack on the citizen's rights. For the government, an AI error may be a statistic, but for the citizen, it can lead to hunger, economic distress, loss of opportunity, and despair.
The irony is that if the existing human negligence, unclear rules, and red tape in government systems are handed over to AI without reform, the problem will not be resolved, but will instead become a digital extension. A flawed system cannot become a perfect system simply by being run by a machine. If old data is discriminatory, AI can perpetuate that discrimination in a new form. If administrative thinking is not sensitive to citizens, technology cannot create sensitivity. Machines may have speed, but they cannot replace a sense of justice, social experience, and human discretion.
Algorithmic bias poses a serious threat to governance. In a diverse society like India, inequalities related to caste, class, gender, region, language, and economic status already exist. If AI begins to make decisions based on data influenced by these inequalities, discrimination against disadvantaged citizens could become even more invisible. Injustice, which used to be visible, can now be hidden behind calculations and technical processes. Excluding someone from a scheme simply because the system did not deem them eligible would not be modern governance, but a new form of digital tyranny.
Governments should also introspect on privacy. Citizens already provide their identity, bank account, health, education, and other details to numerous government systems. Will this information now be used unchecked in the name of AI training and analysis? Is the government's control over citizens' information unlimited? Data security is not just a question of technology, but a question of citizen dignity and freedom. If sensitive information is leaked, responsibility cannot be absolved by simply placing it on a server or software.
The most worrying aspect is the blurring of accountability. A government official may claim the decision was made based on the system's suggestions, the technology company may claim it merely provided the equipment, and the department may claim the process was automated. Ultimately, who should citizens hold accountable? There can be no vacuum of accountability in a democracy. Officials who make decisions by hiding behind machines cannot be allowed to evade administrative responsibility. If AI is used in a government decision, it must also be clear who made the final decision and who is responsible for its consequences.
Transparency in the use of AI in governance should not be a favor but a citizen's right. Citizens should be informed whether AI plays a role in their application, complaint, eligibility, or any other process. What is the basis for the decision, where did the relevant information come from, and what is the avenue for challenging the decision if it is incorrect—answers to these questions must be available. Technical complexity cannot be used as an excuse to evade accountability.
Independent testing and auditing should be mandatory for this purpose. Government AI systems should be regularly evaluated not just for their speed and cost savings, but also for their accuracy, fairness, safety, and social impact. Systems that impact pensions, rations, education, health, employment, and civil rights should have strict oversight and mandatory human approval. In the event of a wrong decision, citizens should have the right to a simple complaint system, reconsideration, and a human hearing.
It is also essential to ensure the security of citizens' personal information in accordance with the Digital Personal Data Protection Act, 2023 in India. Data should be used for limited purposes, unnecessary collection should be avoided, and accountability should be established for security breaches. Government employees should also be trained on the limitations of AI, its false positives, biases, and cyber threats. Digital governance is not merely about purchasing technology; its prudent governance is what truly drives reform.
Another threat is the digital divide. If the government gradually shifts all services to AI-based means, rural citizens, the elderly, the disabled, those with low digital literacy, and linguistically disadvantaged communities could be left behind. Citizens should not be forced to interact with machines to access government services. Technology should facilitate, but not close the door to human assistance.
The government should neither show blind enthusiasm for AI nor reject it outright out of fear. What is needed is a balanced, controlled, and citizen-centric use. High-risk decisions should be mandated with human oversight, independent audits, clear accountability, data security, and effective grievance redressal.
The government's goal should be to solve citizens' problems through technology, not to control them with machines. AI is welcome if it saves citizens from the hassle of running around government offices; but if it leaves them with wrong answers, opaque decisions, and an unaccountable system, it's not an improvement, it's a crisis. No matter how intelligent a machine, democracy demands an even more intelligent system—one where citizens, not convenience, are at the center of decision-making; justice, not calculation; and accountability, not technological excuses.