Print

From Ink and Ledger to Cloud and Cipher: The Bankers' Books Evidence Act, 2026 - A Critical Appraisal of India's Digital Evidentiary Revolution

Published: Sep 18, 2026

By S N Raj, Advocate

I. Introduction: The Anachronism of 1891 and Evolution of the Admissibility of Electronic Evidence

The Bankers' Books Evidence Act, 2026 (New Act) replaces the Bankers' Books Evidence Act, 1891 (Old Act). It is a law finally accepting something every bank customer has known for more than two decades, that a bank today is not a place full of ledgers but a network of servers, cloud storage and digitally signed records, etc. The Bankers' Books Evidence Act, 2026 tries to recalibrate the balance between evidentiary convenience and customer privacy, making evidence easier to use in court while at the same time protecting customer confidentiality.

The 1891 Act belonged to colonial era with quill pens, brass bound ledgers and branch managers who personally knew their depositors. Its central device was elegant in its simplicity, a certified copy of a banking entry could be received in evidence without dragging the original books or their custodians, into court. The rationale was utilitarian, banks could not be paralyzed by incessant summonses and courts could not be expected to verify thousands of folios of double entry book keeping.

Then came the Information Technology Act, 2000 and the amendments of 2002 which patched the definitional gap to a degree as the microfilm, magnetic tape and "any other form of mechanical or electronic data retrieval mechanism" were admitted into the fold and Section 2A created a certification regime for computerized printouts.

The electronic evidence outpaced the 1891 Act, as the said act was drafted for ink and vellum. Indian banking system, however, migrated to core banking software over three decades and Parliament responded by inserting Section 2A into the parent statute and Section 65B into the Evidence Act.

The Supreme Court's jurisprudence on electronic records thereafter, developed in three distinct phases. In the first phase, the Court took a lenient view in State (NCT of Delhi) v. Navjot Sandhu (2005) 11 SCC 600, electronic records were admitted without insisting on a Section 65B(4) certificate as contemplated under the Indian Evidence Act. In the second phase came the decisive correction took place in Anvar P.V. v. P.K. Basheer (2014) - 2014-TIOL-118-SC-MISC, which held that an electronic record is admissible only when accompanied by the certificate under Section 65B(4), overruling Navjot Sandhu (supra) to the contrary extent. In the third phase, a Division Bench in Shafhi Mohammad v. State of Himachal Pradesh (2018) 2 SCC 801 relaxed the requirement where the party tendering the record was not in possession of the device, that relaxation was decisively repudiated by a three-Judge Bench in Arjun Panditrao Khotkar v. Kailash Kushanrao Gorantyal (2020) - 2019-TIOL-302-SC-MISC, which restored and reiterated the dicta of Supreme Court laid down in Anvar P.V (supra). Thus, said certificate is a condition precedent to admissibility, although the Supreme Court softened the rigour by holding, that so long as the trial is not over, the certificate can be directed to be produced at any stage. The wider matrix was completed in State of Maharashtra v. Dr. Praful Desai (2003) 4 SCC 601, which read "evidence" to include modes of electronic transmission even on the updating construction principle and Tomaso Bruno v. State of Uttar Pradesh (2015) 7 SCC 178, which grappled with the proof of digital material in criminal trials.

The 2026 Act, passed by the Lok Sabha on 5th August 2026 and by the Rajya Sabha shortly thereafter, with commencement notified for 1st October 2026, seeks to end this interpretive improvisation by repealing the 1891 Act in its entirety and rebuilding a proper framework on explicitly technology neutral foundations.

II. The Architectural Shift: What the New Act Actually Changes

(a) The expanded definition of "bankers' books"

The 2026 Act's most visible and least controversial, reform is the expansion of the definition of "bankers' books" to embrace records maintained in written or physical form as well as those stored electronically, digitally, virtually or in the cloud, whether kept on-site, off-site or at backup and disaster recovery sites. Whereas the Old 1891 Act (as amended) reached electronic records only through the medium of the "printout", the New 2026 Act recognises the electronic or digital record per se as admissible, valid and legally enforceable evidence, subject to satisfaction of three cumulative conditions:

(i) that the copy is a true copy correctly derived from the original record;

(ii) that no unauthorized change of data is observed or detected, and

(iii) that no tampering with the system or other event compromising integrity and accuracy, has occurred.

(b) Certification and authentication

The 2026 Act introduces standardized certificate formats for certified copies, distinguishing between physical and electronic records and permits authentication by manual, digital or electronic signature. Certification may be effected by the branch head, office head or any other officer nominated by the bank, replacing the 1891 Act's more restrictive identification of certifying personnel. This is a sensible concession to the organizational reality of modern banking, where the person operationally responsible for a record may sit in a Centralized Processing Centre rather than at the branch counter. The reform, however, collapses what the 2000 amendment had carefully separated i.e., the transactional attestation (that the entry is a true copy) and the technical attestation (that the computer system functioned properly, with adequate safeguards). Section 2A of the Old 1891 Act required the latter from "the person in charge of the computer system", who had a technical role. The New 2026 Act, surprisingly and strangely, appears to permit a single generalist officer to certify both a consolidation that trades evidentiary granularity for administrative convenience and one whose consequences, will only be tested in future litigation by adduction of evidence over disputed electronic records.

(c) "Special cause" and the protection of bank officers

Both enactments protect a bank officer from being compelled to produce the bankers' books or to appear as a witness, in a proceeding to which the bank is not a party, save under a court order made for "Special Cause". The Old 1891 Act left that expression undefined, the New 2026 Act supplies a definition of commendable precision. "Special Cause" exists in the following circumstances:

(i) where the accuracy or genuineness of an entry is doubtful ;

(ii) where an event suggests that the regularity or ordinary nature of record keeping has been interrupted; or

(iii) where the bank fails to comply with a court order regarding inspection of its books.

This definitional discipline forecloses the routine summoning of officers in circumstances not covered by the above three clauses.

(d) The Section 11 continuity - Police instead of Court, during investigation

The most contested provision of the Bill was Section 11, which permits orders for the inspection of bankers' books during investigation and inquiry to be made by an officer not below the rank of Superintendent of Police or such other officer as the appropriate Government may specify, in lieu of a court order. Critics read this as a novel transfer of power from the judiciary to the investigating agency. The genealogy, however, is more complicated as Section 8 of the 1891 Act, as amended in 1984, contained a materially identical provision, and it has operated mostly unremarked for over more than four decades. The New 2026 Act's failure to disturb it is therefore, only a continuity and not innovation in its true sense.

(e) Extension to the wider financial ecosystem

Section 4 empowers the Central Government, by notification and subject to such conditions, exceptions or modifications as it may specify, to extend the Act to any entity or class of entities operating in the financial sector. This is the Act's most forward looking and simultaneously, most hazardous provision. The evidentiary presumption of reliability that the statute attaches to a certified bankers' book is tolerable because scheduled commercial banks operate under the Reserve Bank of India's detailed supervisory architecture mandatory audit, prescribed retention periods, capital adequacy and governance standards. To extend the same presumption by executive notification to payment aggregators, fintech lenders or loosely regulated digital platforms is to presume a supervisory rigour that may not exist.

III. The 1891 Act and the 2026 Act: A Comparative Sketch

The comparative picture that emerges is therefore, subtler than the official narrative of modernization suggests. The New 2026 Act's genuine innovations are threefold the direct admissibility of the electronic record, the standardization and liberalization of certification and the notification based extensibility of the regime. Everything else in the New 2026 Act is only a consolidation rather than great revolutionary change.

IV. Inadequacies of the Old Act Remedied by the New

The old definition of a "book" no longer fits reality. The Old 1891 Act imagined a bank record as a physical ledger, courts, in their interpretive wisdom, were therefore, forced to treat a server audit trail as if it were a page in a paper book. The New 2026 Act defines records to include digital and electronic form, so that no such awkward stretching or extended interpretation is now needed.

Under the Old Act Section 2A, a computer record could enter court only through a certified printout. This shifted the fight to the printout, as to whether is this paper an accurate copy? The new Act lets the electronic record itself be ditectly admitted, subject to three simple conditions: (i) it must be a true copy of the original (ii) no unauthorised change must be detected in it and (iii) the system must not have been tampered with. The evidence now speaks for itself rather than through its photocopy.

Cloud and backup records are now clearly covered. Disaster recovery sites and cloud storage were, at best, indirectly covered by the Old Act and courts had to infer that they counted. The New Act names them explicitly, leaving no room or scope for any doubt regarding its contents and admissibility as evidence thereto.

Finally, "Special Cause" has a meaning under the New Act. Under the Old Act, a bank officer could be summoned to court only for "Special Cause", but nobody defined what that meant. Different courts interpreted it differently, producing inconsistent and divergent decisions that lead to chaotic situation and legal debate.

V. The Old Problems the New Act Leaves Untouched

A faithful critical appraisal must record that the New 2026 Act declines to resolve several of the Old 1891 Act regime's most serious pathologies and in one respect, arguably aggravates them.

A person's bank records are a map of their private life which hospitals they visited, whom they donated to, what groups they belong to, what they earn and spend. Yet when these records are demanded in court or during an investigation, nobody has to mask transactions that have nothing to do with the case, nothing stops the disclosed records from being used for other purposes and no remedy exists if the disclosure was wrongful or excessive or misused. The Act does provide three days' notice before an inspection order, but that notice goes to the bank, to protect its operations. The customer, whose records are being opened and scrutinized, has no such privileged rights of privacy at all in the process.

To certify an electronic record, the branch head or nominated officer must swear that the bank's systems had proper safeguards, secure data transfer, network security and cyber resilience. But these matters are known only to the bank's IT department, not to a branch manager who runs daily operations. The Old 1891 Act had a better idea in this respect one person certified that the copy was true and a separate technical officer certified that the computer system worked properly. The New Act merges both into a single generalist certificate. Calling this "simplification" hides the fact that it is a step backward in evidentiary reliability of the deposition in court.

Section 7(2) treats every system a transaction passes through the bank's core system, the payment gateway, the UPI switch, the ATM network, cloud servers as a single unit. This is convenient, but in a cyber fraud case the whole dispute often turns on which link in this chain failed. Treating them as one system is acceptable when nobody questions the record's authenticity, but when authenticity is challenged, the law should require every material system to be identified, and the New Act does not.

Finally, the New Act says nothing about the technical details that actually prove a record has not been altered timestamps, hash values, extraction logs, who entered and who checked the data, change histories or records of data migration between systems. These are exactly the details that decide disputes about backdated entries, unauthorized transactions and "Wilful Default" findings. The legislature chose not to require them, one can only hope that courts, under the Bharatiya Sakshya Adhiniyam, 2023, will read them in.

VI. How the New Act Will Actually Work: Four Practical Consequences

First, electronic records can now be used directly in court, with standard certificates and digital signatures. This removes the costly process of making printouts and the arguments about whether the printout is accurate. Recovery suits, cheque-bounce cases and enforcement matters should therefore, move noticeably faster.

Secondly, as bank records become easier to produce, banks gain an even bigger advantage over borrowers. A digitally certified statement will carry heavy initial weight in court, small borrowers, pensioners and those not comfortable with technology will find it extremely hard to challenge these certificates without a lawyer. The New Act does not say that a bank entry is only initial evidence, nor does it oblige the bank to show the underlying audit trail when an entry is specifically challenged. The danger is clear making procedure faster must not end up making justice harder for the weaker party.

Thirdly, Section 4 lets the Government extend the New Act to NBFCs, payment companies and fintechs. If this is done carefully, with proper regulatory conditions, it could improve record keeping across the industry, if it is done carelessly, it would result in adverse consequences, including breaches of data privilege norms and the confidentiality obligations of NBFCs.

Fourthly, during an investigation, a police officer can order the production of banking records. Parliament thus had the perfect opportunity, while rewriting a 135 year old law, to require a judge's prior approval before a citizen's bank records are handed over to the police but unfortunately, this has not been done in the New Act. The New Act endeavors to improve efficiency everywhere, but distributes the benefits unevenly as on one hand banks and investigators gain immediately, while on the other hand customers must wait for courts to provide the protection the statute withholds.

VII. Conclusion

The Bankers' Books Evidence Act, 2026 fixes a problem that should have been fixed atleast three decades ago. It gives legal recognition to cloud storage, digital records and backup systems, it standardizes certificates and limits unnecessary summonses and it rightly extends banking evidence rules to the wider financial world. These are genuine achievements and the New Act is also now in consonance with developments under the Information Technology Act, the jurisprudence on electronic evidence laid down by Supreme Court from time to time and the Bharatiya Sakshya Adhiniyam, 2023.

However, the New Act modernizes the technology without modernizing the protections due to the customer that is only half the job done. When someone's bank records are demanded, that person has no say or right in the matter. The branch manager still certifies the security of computer systems he has never seen. The police officer who orders the records is the same officer whose investigation will use them. The Old 1891 Act survived for 135 years not because it was well written, but because courts applied it with due care, circumspection and patience. The New Act gives courts better tools, but whether it protects ordinary citizens remains to be seen as banking records can be completely genuine and still be deeply invasive of someone's privacy.

The crux is that the ink has finally dried on the ledgers and under the Bankers' Books Evidence Act, 2026, the servers are humming. The law has eventually caught up with where the financial accounts, data and digital records presently live, but the huge challenge to the data privacy of bank customers still remains the area of grave concern.

[The views expressed are strictly personal.]