In Jeethu Joseph’s Malayalam thriller Drishyam, Georgekutty, played by Mohanlal, protects his family by constructing a carefully organised account of events. He anticipates questions, arranges details and understands how people interpret the evidence before them. The Hindi adaptation, starring Ajay Devgn, brought the same unsettling idea to a wider audience. The film is fiction, but its central insight is relevant to modern political communication: people do not always judge evidence independently. They often interpret it through a story they already believe.
In the digital age, political narratives can be repeated across thousands of accounts, amplified by recommendation systems and reinforced through videos, headlines, commentary and online discussions. Eventually, the sheer frequency of a claim may make it seem proven. That is the Drishyam effect in public life. A carefully constructed narrative becomes so familiar that questioning it appears unreasonable, even when the evidence supporting it remains incomplete.
The danger is particularly serious when political actors use this process to undermine public confidence in democratic institutions. An allegation against an elected government is not automatically false. Nor is it automatically true. Yet, in the attention economy, the distinction between an allegation and a fact can disappear beneath the weight of repetition. A landmark study published in Nature provides an important scientific starting point for understanding this problem.
What the nature study reveals
The study, The Political Effects of X’s Feed Algorithm, published online on February 18, 2026, examined the influence of an algorithmic social media feed on political attitudes and behaviour. The field experiment involved 4,965 active X users in the United States over seven weeks. Researchers compared users exposed to the platform’s algorithmic feed with users receiving a chronological feed. A chronological feed presents posts in sequence, while an algorithmic feed ranks content according to a system designed to determine what users are more likely to engage with.
The researchers found that switching to the algorithmic feed shifted some political attitudes in a more conservative direction. It also changed the political content participants encountered and increased their following of conservative activist accounts. Some changes in following behaviour persisted after participants returned to the chronological feed. The findings were not unlimited. The researchers did not find significant changes in self-reported political affiliation or affective polarisation. The experiment also did not directly establish that repeated falsehoods become believable simply because an algorithm presents them repeatedly.
These qualifications matter. The study should not be used to claim that every algorithm is designed to produce a particular political outcome, or that citizens have no control over their beliefs. Its important finding is narrower and still consequential: the design of a social media feed can change the information people encounter and influence some political attitudes.
The research was conducted in the United States, and its findings cannot automatically be generalised to Bharat’s multilingual, politically diverse society. Nevertheless, it raises a question relevant to every democracy: when digital platforms determine which political messages receive greater visibility, how does that affect the way citizens understand public affairs? The algorithm does not need to invent a false story. It may be enough to give an existing allegation repeated visibility.
When repetition becomes a substitute for proof
Psychological research on the illusory truth effect has shown that repetition can increase the perceived truth of statements. Familiarity can influence judgement even when a person does not consciously recognise its effect. In politics, this creates an opportunity for manipulation. An allegation appears in one post, is repeated by commentators, becomes the subject of a video and is then cited by other accounts as evidence that the allegation is widely believed.
But the number of times a claim is repeated is not the number of independent pieces of evidence supporting it. A thousand posts repeating the same unverified allegation may still amount to just one unverified allegation. Political communication becomes especially misleading when an emotionally powerful conclusion is presented first and supporting details are selected afterwards. A photograph without context, an old video presented as a new event or a complicated administrative figure reduced to a dramatic slogan can become the foundation of a political narrative.
Once such a narrative becomes established within an online community, corrections may struggle to reach the same audience. People may remember the accusation but forget the explanation. They may recall that an institution was accused of wrongdoing without remembering whether the accusation was substantiated. The process can damage trust even when the original claim is later contradicted.
The politics of manufactured suspicion
Political parties have every right to challenge governments. Opposition parties are essential to democracy because they expose failures, question decisions and offer alternatives. Investigative journalism, public protest and scrutiny of constitutional institutions must remain protected. But opposition is not exempt from the obligation to establish facts.
When a party converts an allegation into a declaration of proven criminality before the supporting evidence has been independently established, it risks misleading the public. When a political slogan replaces the detailed examination of records, the slogan can become more influential than the facts. The ruling party and its supporters must meet the same standard. No government should be protected from legitimate scrutiny merely because an allegation is politically inconvenient. But neither should an institution be condemned because a claim against it has been repeated often enough.
The issue is not whether political communication is partisan. It inevitably is. The issue is whether partisan communication respects the difference between evidence, inference and accusation. This distinction is especially important when the target is the legitimacy of an election. Electoral confidence is not strengthened by demanding unquestioning faith in the Election Commission. It is strengthened by examining evidence, applying the law and refusing to treat unsupported conclusions as facts.
SIR and the attempt to turn voter-list revision into proof of electoral fraud
The controversy over the Special Intensive Revision (SIR) of electoral rolls in Bihar illustrates how a complicated administrative exercise can become a political narrative. The Election Commission announced the Bihar SIR in June 2025. The exercise involved examining electoral records, identifying ineligible entries and providing opportunities for eligible electors to be included. The final roll published on September 30, 2025 contained approximately 7.42 crore electors, compared with 7.89 crore on June 24.
According to the Commission’s published figures, 65 lakh names were removed before the draft roll was published on August 1. A further 3.66 lakh entries were removed between the draft and final rolls, while 21.53 lakh eligible electors were added during the process through Form 6. These figures describe different stages of the exercise and should not be conflated. The central principle is straightforward: an electoral roll must include eligible voters and exclude ineligible entries. Duplicate registrations, deceased electors and people who have permanently shifted must be dealt with according to the applicable rules. At the same time, no eligible citizen should be wrongly deprived of the right to vote.
The existence of deletions is not, by itself, proof of vote theft. A deletion figure must be examined by category and against the underlying records. Equally, the addition of eligible voters does not prove that every deletion was correct. Each question must be assessed on its own evidence. The Election Commission has an established statutory role in maintaining electoral rolls. Its decisions should be assessed under the applicable law and procedures, not through a presumption that every revision is politically motivated. A political party alleging organised electoral fraud must provide evidence connecting specific entries or actions to the alleged wrongdoing.
Form 6 is particularly relevant to this discussion. It is the prescribed application for the inclusion of an eligible elector’s name in the electoral roll. The Election Commission’s voter services portal provides access to Form 6 and the associated registration process. I have personally seen a name removed from the electoral roll and subsequently restored after Form 6 was filed. That experience does not prove that every case will be resolved in the same way, but it illustrates an important point: a missing name is a problem to be addressed through the prescribed process, not automatic proof that an election was stolen.
Eligible voters should check their registration before an election and use the available procedures to seek inclusion or correction. Electoral authorities must process such applications fairly and lawfully. Political parties should help citizens identify genuine errors and pursue remedies rather than turn every discrepancy into evidence of a national conspiracy. The appropriate response to a wrongful deletion is correction and accountability. The appropriate response to a duplicate or ineligible entry is lawful removal. Neither situation requires the public to abandon evidence-based judgement.
Three claims that deserve closer scrutiny
1. The claim that 1,00,250 votes were stolen in Mahadevapura
In August 2025, Congress leader Rahul Gandhi alleged that 1,00,250 votes had been stolen in the Mahadevapura Assembly segment of Bengaluru Central during the 2024 Lok Sabha election. He presented figures relating to alleged duplicate voters, questionable addresses, multiple registrations, photographs and misuse of Form 6. The distinction between identifying anomalies and proving theft is critical. A duplicate-looking entry, an unusual address or a disputed registration requires verification. It does not automatically establish that a fraudulent vote was cast, that the same person voted multiple times, or that election officials and the BJP conspired to steal the election.
The Karnataka Chief Electoral Officer asked Gandhi to submit a declaration under oath and provide the names of electors allegedly involved so that the allegations could be examined through the prescribed process. That request went to the heart of the matter. If a political party possesses evidence of criminal electoral fraud, it should identify the specific entries, submit the supporting material and enable the competent authorities to investigate. A press conference is not a substitute for that evidentiary process.
The figure of 1,00,250 should therefore not be repeated as though it were an officially established count of stolen votes. It was a figure advanced by Congress based on its own analysis. The allegation requires proof of what the anomalies actually represented and whether they translated into unlawful voting or affected the result.
2. The broader claim that the 2024 general election was stolen
Claims about electoral manipulation have circulated through political speeches and social media, including allegations concerning electronic voting machines and voter registration. But suspicion about a system is not proof that the system was manipulated in a particular election. On 26 April 2024, the Supreme Court rejected petitions seeking 100 per cent verification of VVPAT slips against EVM counts and a return to paper ballots. The Court examined the safeguards governing the EVM system and did not find the material before it sufficient to justify the requested changes. It also provided a mechanism for post-result verification of the burnt memory or microcontroller of a limited number of EVMs on a candidate’s written request.
That judgment does not mean that every conceivable allegation about every election is impossible. It does mean that a sweeping claim of systemic fraud cannot be established simply by repeating doubts that have already been examined by the courts. Specific evidence must be produced and assessed under the applicable law.
3. The claim that Bihar’s later SIR proves the Prime Minister was illegally elected in 2024
The chronology alone does not support this conclusion. The Bihar SIR was announced in June 2025, after the 2024 general election. A subsequent electoral-roll revision does not automatically establish that the earlier election was illegal. To substantiate such a claim, a complainant would need evidence tied to the relevant 2024 electoral rolls and constituencies, a demonstration of the alleged unlawful inclusion or exclusion, and a legally sustainable basis for connecting those facts to the election result.
The Supreme Court did ask the Election Commission for clarification regarding the approximately 3.66 lakh names removed between Bihar’s draft and final rolls in October 2025. That was a matter requiring clarification; it was not, by itself, a judicial finding that the 2024 general election had been stolen. There is a fundamental difference between questioning an administrative process and proving that an earlier election was illegally decided. The first may be entirely legitimate. The second requires evidence that establishes the specific legal claim.
These examples reveal the danger of political communication that jumps from an anomaly to a conclusion, from a question to an accusation, and from an accusation to a declaration of guilt. The more dramatic the claim, the greater the need for precise evidence.
Bharat must not confuse criticism with proof
Bharat’s democratic institutions should be open to scrutiny. The Election Commission must follow the law, maintain accurate rolls and provide remedies when errors occur. Courts must examine genuine disputes. Journalists must investigate credible allegations. Opposition parties must remain free to question the government. But these principles do not require the public to accept every accusation against a constitutional institution as true.
The same standard must apply to all political parties. If the BJP makes a claim about its opponents, it must support that claim with evidence. If Congress or another opposition party alleges electoral fraud, it must do the same. Political advantage cannot determine the standard of proof. Social media algorithms complicate this task because they can reward material that attracts attention. A dramatic allegation may travel further than a careful explanation. A correction may reach fewer people than the original accusation. The resulting imbalance can leave citizens with a strong impression of wrongdoing without a clear understanding of whether the allegation was substantiated.
This is why transparency from platforms matters. It is also why political parties and news organisations must distinguish between verified facts and claims that remain unproven. Citizens, meanwhile, should ask whether a widely shared allegation is supported by original records, independent reporting or a finding by a competent authority. The objective is not to discourage criticism. It is to make criticism more credible by insisting on evidence.
Beyond the Drishyam effect
In Drishyam, a carefully constructed story shapes the way events are interpreted. In political life, repeated narratives can similarly shape perception, especially when algorithms repeatedly deliver familiar claims to receptive audiences. But a democratic society cannot afford to treat repetition as proof. An allegation does not become true because it trends. An anomaly does not automatically prove criminality. A later revision of electoral records does not, by itself, invalidate an earlier election.
The Nature study warns that algorithmic feeds can influence political exposure and some attitudes. It does not establish that every political narrative is engineered or that every viral allegation is false. Its lesson is that the information environment matters, and that citizens must remain alert to the difference between what they see frequently and what the evidence actually establishes.
For Bharat, democratic resilience requires confidence in lawful institutions alongside a willingness to correct genuine errors. It requires opposition parties that challenge the government with evidence, governments that answer legitimate questions, journalists who verify claims, and citizens who refuse to outsource their judgement to a trending hashtag. The strongest defence against misinformation is not blind trust in authority. Nor is it automatic suspicion of every institution. It is a consistent standard of evidence applied regardless of who benefits from the conclusion. A story may be repeated a thousand times. It may be emotionally satisfying, politically useful and algorithmically amplified. None of those qualities makes it true.
Bharat’s strongest defence against algorithmic distortion is a citizen who understands that hearing a story a thousand times is not the same as having it proved even once.
