Author: Anjali Bharti, a practicing Advocate at Patna High Court, Patna Bihar.
ABSTRACT
The authentication of electronic evidence is becoming more pertinent because of the increasing reliance on digital records in legal proceedings. The Bharatiya Sakshya Adhiniyam, 2023 (BSA) has replaced the Indian Evidence Act, 1872, which offers a new regime with respect to the admissibility of electronic records under Section 63 of the BSA. The provision is continued using the approach of certification, similar to that found in Section 65B, but adds other requirements, such as the mandated Part A certificate and the disclosure of cryptographic hash values under Part B. Actions like these are intended to improve the trustworthiness of digital evidence, but when the material is not just replaced by artificial intelligence, or can be modified or edited, it is doubted if these measures will be viable.
This paper examines the effectiveness of the protections in Section 63 to preserve the authenticity of electronic evidence in the current electronic environment. The report emphasises the difference between the integrity of a digital file and the veracity of the information in the file. The study explores the legal basis, the relevant case law, the novel technological problems presented by synthetic media and deepfakes, and comparative methodologies to digital evidence authentication. The research methodology is mainly doctrinal with some comparative analysis and empirical and institutional materials.
The research reveals that Section 63 offers a more detailed mechanism for recording and retaining electronic evidence, however, certification and hash value verification are insufficient to verify the origin, truthfulness or factual authenticity of digital content. This restriction is especially relevant when it comes to content created or manipulated by AI that can look like real evidence. The study thus suggests a multi-layered and differentiated approach to authentication by implementing statutory certification and digital forensic examination, as well as authentication mechanisms based on provenance and increased scrutiny of disputed and/or sensitive synthesized evidence.
Keywords: Bharatiya Sakshya Adhiniyam, 2023; Electronic Evidence; AI-Generated Evidence; Deepfakes; Digital Forensics; Authentication; Hash Value; Content Provenance.
- INTRODUCTION
Digital technology is ubiquitous today, and this has inevitably had its way into the legal process. Information from emails, WhatsApp conversations, CCTV, photographs, call records, social media etc. may be required by modern courts as well as computer and/or mobile phone information. Electronic records, in comparison to traditional documents, can be copied, transferred and modified easily. The value of evidence is, therefore, not only related to the contents of the document, but also to the documentation of the origin, integrity and preservation process of the document.
With the rapid growth of generative AI, the problem has become even tougher. AI tools are now capable of producing realistic photos, videos and audio recordings that appear to be of real people or events. That means that people can create deepfakes and other fake media that can pretend to be evidence of an event that did not happen. This turns the typical question of electronic authentication on its head. The question “Has there been any change in a digital file since it was created?” is no longer good enough. Just as a court may have to determine whether the content was genuine when it was produced, whether it originated from the person or even the story depicts, and whether it can be confidently linked to the individual or event, it will also have to consider whether it is reliable.
Bharatiya Sakshya Adhiniyam, 2023 (BSA) has completely replaced the Indian Evidence Act, 1872, marking a major change in the evidence laws of India. The new Section 63 in the BSA retains the same certificate-based approach that developed under Section 65B of the previous Act with the addition of covering electronic records. Concurrently, it offers a finer-grained certification mechanism, consisting of Part A and Part B of the required certificate, and information about electronic devices and cryptographic hash values. These changes appear to suggest a way of making it more systematic to handle and verify digital evidence.
The previous judicial rulings, Anvar P.V. v. P.K. Basheer and Arjun Panditrao Khotkar v. Kailash Kushanrao Gorantyal, had established important precedents on the issue of certification and admissibility of electronic records. However, the decisions were taken in a different technological context that were not fully prepared to accommodate the current state of generative AI and synthetic media.
Most of the discussion about Section 63 thus far has been procedural changes, the certificate requirements and a comparison with Section 65B. There has been less focus on how these safeguards can be effective when they have to protect the authenticity of content that might also be artificially generated. For instance, a hash value can be used in a way to indicate that a particular file has not changed since it was hashed, but by itself cannot indicate that the contents are from a real event.
In this backdrop, the present study examines the effectiveness of Section 63 to address the contemporary issues of electronic evidence. It explores the connection between integrity and authenticity, explores the utility and limitations of a certification and hash value, and talks about the evidentiary problems that could arise with content created by AI, among others, and how the legal landscape in India may be strengthened by other forensics and provenance-based safeguards.
- LITERATURE REVIEW
The existing scholarship in India on electronic evidence has been developed primarily in terms of Section 65B of the Indian Evidence Act, 1872. The earlier legal battle had been regarding the method by which electronic records were admitted and the importance of the certificate needed by the provision. The Supreme Court in its two judgments in Anvar P.V. v. P.K. Basheer and Arjun Panditrao Khotkar v. Kailash Kushanrao Gorantyal were useful in clarifying the evidentiary requirement for electronic records. These rulings established that electronic evidence is not admissible in exactly the same manner as traditional documentary evidence and that statutory safeguards will play a significant role in the admission of electronic evidence.
Legal commentary of late has concentrated on Section 63 and its new certification requirements with the enactment of the Bharatiya Sakshya Adhiniyam, 2023. The focus of the discussions has been on the differences between Section 63 and Section 65B, how to use Part A and Part B of the certificate, identification of electronic devices, and the need for providing hash values. While this literature can provide an understanding of the changes in the legislation, a significant portion of it still remains devoted to understanding the new provision, but not to testing it against the new types of digital manipulation.
At the same time, research in digital forensics has focused on the importance of maintaining the integrity and chain of custody of digital material. There are also findings from research across various countries that have been found to be a problem in the field of deepfakes and synthetic media: digital content may be replicable at the time and so may appear as an actual event but actually not be.
The research gap is to link the discussion of these two topics. Little analysis has been conducted on how much can be achieved by the safeguards in Section 63, such as certification and hash-value verification, in addressing the authenticity problems posed by AI-generated and manipulated content. The purpose of the present study is to explore potential differences between a digital record that has not been altered and content that is actually authentic, using the new framework. It also questions whether greater safeguards are required to ensure that electronic evidence law is responsive to the evolving digital world, such as more forensic and provenance-based protections.
- 3. RESEARCH METHODOLOGY
3.1 Research Design
The research method used is doctrinal legal research, combined with comparative and empirical approaches. Using the doctrinal method of research, the statutory provisions and judicial decisions pertaining to electronic evidence in India are being analysed. Comparative analysis is used to find approaches that are relevant to other jurisdictions and empirical material to understand the context of the technology and evidence.
3.2 Sources of Data
The primary sources are the Bharatiya Sakshya Adhiniyam, 2023, the previously existing Indian Evidence Act, 1872, judicial rulings and pertinent governmental documents. Secondary sources consist of scholarly articles, books, legal commentaries, literature, and reports from recognized digital-forensics organizations on the Internet. Statements related to the use of digital evidence, cybercrime and synthetic media are backed up by the available government data, international databases and industry reports.
3.3 Scope
The research is focused on Section 63 of the BSA specifically the certification requirements, Part A of the prescribed certificate and cryptographic hash values, part B of the prescribed certificate. It takes into account these requirements in the context of authenticity and integrity of electronic evidence, including AI-generated and manipulated digital evidence. The evolution of the law is discussed in the context of Section 65B and the case law.
3.4 Limitations
Section 63 is a relatively new legislation and the interpretation of the section by the judiciary is still limited. Also, AI and digital-forensic technologies are rapidly developing. These factors limit the availability of long-term empirical research and there is also less finalised judicial doctrine. The research thus depends on the most up-to-date and authoritative sources available at the time.
3.5 Analytical Framework
The study assesses the effectiveness of Section 63, which can be broken down into four related concepts: admissibility, integrity, authenticity and provenance. This framework is used to see if the existing certification mechanism can certify a digital record as unaltered, or certify the source of the record as well as the authenticity of the content.
- DISCUSSION AND ANALYSIS
4.1 From Section 65B to Section 63
In the past decade, the way electronic evidence is being treated in India has undergone a dramatic transformation. Under the provisions of Section 65B of the Indian Evidence Act, 1872, the electronic records also cannot be “proved” in the same way as the ordinary documents. In Anvar P.V. v. P.K. Basheer the Supreme Court has given importance to the statutory certificate for proving electronic records. This was later clarified in Arjun Panditrao Khotkar v. Kailash Kushanrao Gorantyal where the Court in this case reiterated the importance of the certificate but also addressed the practical problems of obtaining the certificate.
Section 63 of Bharatiya Sakshya Adhiniyam, 2023 uses this approach, but with a more specific set of rules for electronic records. The prescribed certificate requires details about the electronic record, device or source and other technical details. The introduction of Part A and Part B and the need to expose a cryptographic hash value are signs of attempts to make the process of verification more systematic.
These changes, while helpful, do pose a larger question. Is the authenticity of the contents of an electronic record established by the mere fact of being properly produced and preserved? It is not always OK to say yes. That difference is particularly germane if the content is synthetically created.
4.2 What a Hash Value Can—and Cannot—Prove
An important technical detail of the new framework is the addition of a cryptographic hash value. A hash is created from digital information. It can be used to check if that information has changed since that time. Usually, a file will have a different hash if even a small change has been made to it. Once this is done, hashing can be useful to establish the integrity of a given digital file.
But integrity and authenticity are two distinct attributes. When hashes match, it can be proven that the file displayed at a later time is identical to the file hashed at an earlier time. It does not assert who created the file and whether the hardware or other device that captured the event did so accurately, or whether the event depicted or described in the file did in fact happen.
For example, a video produced by AI can be hashed without any modification. Later, when the same file is created before a court, it is also possible that the hash can be used to help prove that the file hasn’t been altered. It cannot establish that the video is an actual event. The hash requirement is thus a technical means to ensure that the content will not be changed at a later stage, but cannot be interpreted as proof of the accuracy of the content.
4.3 Research findings and the New Authentication Problem
With the development of generative AI, the difference has become even more significant. Common issues with electronic evidence include whether an audio or video recording had been edited, if a photograph was altered, or if a digital file had been altered after being collected. Another challenge is posed by AI-generated content, as it can be wholly synthetic from the moment it is created.
The picture could now represent someone who was never in that location. A voice can be generated to sound like a particular individual without that person making the statement. Videos can also be edited to produce what looks like a genuine event. In these cases, it may not be obvious from looking at the final file that it has been changed.
The concept of integrity therefore needs to be broadened to take into consideration authenticity and provenance. The court may wish to know the source and origin of the material, how it is created, if it has been edited with editing software or generative systems, and if there is reliable information that links the material to the event that it claims to represent.
A pertinent technological advancement is the Coalition for Content Provenance and Authenticity (C2PA). It offers standards to capture information about the origin and history of digital content using cryptographically verifiable provenance information. This kind of data may be utilized to distinguish the way a resource was made or later altered.
However, don’t overlook the importance of judicial appraisal as an alternative to C2PA. While provenance can be helpful to determine the history of an asset, it is notdefinitive of the factually truth of the contents. Its usefulness is proving another source of information in authentication.
4.4 Digital Forensics Principles and Concepts
As electronic evidence evolves, so does the significance of digital-forensic examination. When the authenticity of a record is seriously in doubt, the court may require more information than is contained in the statutory certificate. Original device, file properties, chain of custody, and extraction techniques could be used to establish if evidence was altered or if the source of the evidence is believable.
This is not to say that all WhatsApp messages, photos or CCTV footage need to be strongly subjected to an in-depth forensic examination. If such a mandate were to be imposed upon the justice system, it would be needlessly burdensome and costly. What would be more practical is to have different levels of scrutiny depending on circumstances of the evidence.
The Section 63 procedure can be used where the source is clear and there is no significant challenge. Forensic examination may be considered if a legitimate question is raised by the opposing party about alteration or origin. If there are particular indicators of deepfake or AI influence, then a more thorough technical evaluation would be appropriate.
4.5 Overall Assessment
This is a significant step in India’s evolving jurisprudence on electronic evidence, Section 63. It has the potential to enhance documentation and preservation of digital records with its certification and hashing mechanism. However, it doesn’t answer all the questions users might have about the authenticity of digital content.
The key difference, very simply, is that a file can be not only intact but not authentic. Even if the hash of the AI-generated video matches the original video, it is still considered AI-generated. The evidentiary context must therefore include evidence as to how and when the record was changed, as well as who created it, its relationship to the event it purports to represent, and the source of the evidence.
A multi-layered approach would be most appropriate. Section 63 can offer the basic statutory context, and any forensic evidence, chain of custody and provenance information can be enhanced where necessary because of the circumstances. This would enable courts to handle normal electronic evidence properly and still be able to handle complex forms of synthetic evidence.
- FINDINGS
The review of Section 63 reveals that BSA has paved the way for the electronic evidence collection process to be more structured. Part A and Part B of the certificate and the need to provide a hash-value are helpful in understanding the source and preservation of a digital record. These protections can help a court determine if the court-approved process has been completed.
The study also pinpoints the significant drawback of heavily depending on file integrity. Completely genuine digital content can be inserted into a digital file and added back. A hash cannot prove that the file hasn’t been altered since it was hashed, it cannot prove that the underlying event occurred, and it cannot prove that the material accurately portrays the event.
This issue becomes more problematic with AI-generated photos, audio and videos. When the recorded item itself is synthetic, the traditional problem of detecting changes to the existing record is no longer adequate.
Based on the analysis, it is then concluded that Section 63 is the beginning, not the end, of the authentication process. If the authenticity of evidence is in doubt, further examination of metadata, chain of custody, evidence device, forensic indicators and provenance could be required.
The best way to do this is to take a proportionate approach; normal electronic records can be processed in the same way as the statutory process, whilst disputed and/or potentially AI-generated records may warrant more technical review. This would make digital evidence more reliable without unnecessary complications in all cases.
- RECOMMENDATIONS
The results do not suggest that Section 63 needs to be replaced entirely. Rather, the provision needs to be supported by safeguards that respond to the changing nature of digital evidence.
A risk-based authentication system would be a good place to begin. Not all electronic records are the same in terms of potential tampering. As long as the source is clear and the record is not seriously disputed, the requirements of Section 63, including certification and hash, should normally be enough. If the authenticity of a record is truly in dispute, then the court should be able to consider having a different forensic examination done. Any evidence that appears to be generated by AI or heavily manipulated should be scrutinized more closely.
There is also the need for uniformity in the processes of collection and preservation of electronic evidence. Investigating agencies and forensic laboratories must adhere to clear standards for preservation of original files, documentation of file origin, hash values and chain of custody. The consistency of the investigation would minimize controversy about the preservation of digital evidence later on.
Also, the number of qualified digital-forensic experts should be increased. With the increasing sophistication of electronic evidence, court may seek expert assistance to identify manipulation that can’t be found by a normal examination of the evidence. Training should include new types of synthetic media such as images, audio and video created using artificial intelligence.
Another form of protection can be digital provenance. C2PA or similar standards might be used to report information about the origin or history of digital content. Provenance should be used as a guide, however, and should not be used to assume that the content is factually true.
Lastly, more definite guidance from the courts is needed to differentiate between admissibility and integrity/authenticity. A certificate issued under Section 63 should provide the statutory basis for electronic evidence, but the court should not be barred from seeking further evidence of genuineness where the genuineness of the evidence is reasonably challenged.
These measures would form a proportionate multi-layered system that would serve to safeguard the evidentiary value of authentic digital content without unnecessary technical demands in all cases.
- CONCLUSION
Bharatiya Sakshya Adhiniyam 2023 has taken a massive step towards making Indian law of evidence more applicable in the digital sphere of communication. Section 63 elaborates on the process for presenting electronic records, such as the use of cryptographic hash values, as well as the requirements for certification. Such measures can be helpful in determining the creation of a digital record, and the integrity of that record.
It is important to note, however, that an unchangeable file is not necessarily a true file. With the development of generative AI, photographs, recordings and videos can be created or manipulated to represent events that never took place. A hash can be used to ensure the integrity of such a file, but cannot be used to determine if its contents are reliable. For authenticity to be achieved, digital integrity is not the only requirement.
The study concludes that Section 63 offers a significant statutory underpinning, but is insufficient to solve all the challenges offered by the current synthetic media. If the authenticity of the electronic evidence is seriously questioned, forensic investigation, proper chain of custody and sound information regarding the origin and history of the contents can be required.
Indian Evidence Law should therefore shift to a pragmatic and sensible approach to authentication. There is no need for additional technical procedures to burden ordinary digital records, but closer examination of evidence that has a real risk of manipulation by AI. This would maintain the viability of Section 63 while enabling the law to adjust to newer technologies that evolve even more rapidly than the conventional ways of authentication.
