The Election Commission of India has declined to disclose the name and version of the software it used to identify errors and anomalies in electoral rolls during the Special Intensive Revision, saying the information relates to commercial confidence, trade secrets and intellectual property and is therefore exempt from disclosure under the Right to Information Act.
The reply, dated October 16, came in response to a Right to Information application filed on July 27 by social activist M A Mujeeb Ayyub from Telangana. Ayyub had sought details of the technology used in the voter-roll revision process, including the software, application, algorithm or artificial intelligence platform used to detect errors, anomalies or inconsistencies in electoral rolls during the Special Intensive Revision, known as SIR.
In its response, the poll body said no new software had been developed or procured for the SIR exercise. It added that Electoral Registration Officers decide whether to include or exclude an elector based on facts and records, a point the Commission returned to repeatedly in answering different parts of the application.
Ayyub had also asked whether the system automatically flagged voter records for verification, correction or deletion and, if so, what process was followed after such records were identified.
He further sought to establish whether artificial intelligence, machine learning, predictive analytics or automated decision-support systems were used at any stage of the exercise, and what safeguards existed against false positives, incorrect identification of anomalies or wrongful flagging of genuine voters.
On those questions, the Commission declined to give substantive answers. It said the information sought about the software's name and version was exempt from disclosure under Section 8(1)(d) of the Right to Information Act, 2005, which covers information relating to commercial confidence, trade secrets or intellectual property, the disclosure of which would harm the competitive position of a third party.
On the questions about how flagged records were handled and whether automated systems were involved, the Commission said the information sought was a clarification which does not come under the definition of information in Section 2(f) of the Act.
The Commission also invoked Section 8(1)(g) of the Act in response to questions about the review and approval of flagged records, and referred to Sections 13B(1) and 13B(2) of the Representation of the People Act, 1950, which assign responsibility for preparing and revising electoral rolls to the Electoral Registration Officer concerned.
The combined effect of the replies is that the Commission has neither confirmed nor denied the use of artificial intelligence or automated systems in identifying irregularities in voter records, and has not disclosed what software, if any, performs that function.
An Election Commission official said software and algorithm details were conventionally not shared under Right to Information requests. The official pointed out that even manufacturers of electronic voting machines, including Bharat Electronics Limited and Electronics Corporation of India Limited, had earlier refused to share source codes while replying to similar queries, treating such details as proprietary.
The disclosure has drawn sharp political criticism. The Congress has accused the Commission of withholding information about its ECINET software.
Party leader Jairam Ramesh questioned why the Election Commission had not itself publicly disclosed that Tata Consultancy Services developed the software, and raised questions about the use of artificial intelligence during the Special Intensive Revision and electoral roll revisions, including the safeguards in place to prevent errors and ensure transparency.
The controversy touches on a sensitive aspect of the revision exercise. The Special Intensive Revision involves a house-to-house verification of electoral rolls, and the accuracy of the resulting voter lists has become a subject of intense political debate across the country.
Opposition parties have repeatedly alleged wrongful deletions and errors in the rolls, while the Commission has maintained that inclusion and exclusion decisions rest with Electoral Registration Officers acting on facts and records, not with any automated system.
Transparency advocates argue that when technology is used to flag citizens' voter records for verification or deletion, the public is entitled to know what that technology is, how it works and what checks exist against mistakes. The Commission's position, as reflected in its reply, is that the software details are commercially sensitive and that the statutory framework already places decision-making responsibility on human officers.
The legal provisions cited by the Commission are central to how this dispute will be understood. Section 8(1)(d) of the Right to Information Act allows public authorities to withhold information whose disclosure would harm the competitive position of a third party, unless the competent authority is satisfied that a larger public interest warrants disclosure.
Section 2(f) defines information as material in any form, including records, documents and opinions, but the Commission's view is that a request seeking an explanation or clarification, rather than an existing record, falls outside that definition. Applicants who disagree with such a reading can pursue first appeals within the public authority and second appeals before the information commission, and the dispute over the SIR software may yet travel that route.
For now, the practical outcome is a gap in public knowledge. The Commission has said no new software was developed or procured for the revision exercise, which suggests that any system in use predates the exercise, but it has not named that system or its version. It has said officers decide inclusion and exclusion on facts and records, but it has not described the workflow by which records are flagged for those officers in the first place.
And it has not answered whether artificial intelligence, machine learning or predictive analytics played any role at any stage.
That gap is likely to keep the issue alive. The revision of electoral rolls affects the voting rights of millions of people, and questions about the technology behind it sit at the intersection of administrative transparency, data governance and electoral integrity. As long as the Commission declines to name the software or describe its safeguards, critics will continue to ask why a process that determines who gets to vote should rely on tools the public cannot examine.
The episode also highlights a broader tension in how public authorities handle requests about technology. Governments increasingly rely on software developed by private vendors or on platforms whose inner workings are treated as proprietary. When citizens use transparency laws to ask about those tools, authorities must balance commercial confidentiality against the public interest in understanding decisions that affect fundamental rights.
The Commission's reply in this case has come down firmly on the side of confidentiality, and it is that choice, as much as the software itself, that is now under scrutiny.
(Reporting based on The Times of India and The Economic Times.)