Electronic Evidence: A Way Forward

Court cases, whether civil or criminal, are how we ask the legal system to decide contested questions of fact. Both sides give their versions of what happened by leading evidence of the many details that made up the event. Witnesses are called, documents and other materials filed, all to show that one version was more probable. A court then examines these materials to decide which version of events was probably (or most possibly) true. 

As trials are attempts at recreating what people did, it is natural to imagine that a major way of doing this is through mobile phones, computers, and social media. Not only do many of us create an almost contemporary record of their daily life on Facebook etc., but cellphones are also sophisticated tracking devices that can help locate where someone was on any given day. At the same time, the likelihood of fake accounts and tampering with this material is also quite high, making it difficult to believe all that we do end up seeing.

This throws up the following dilemma: Should it be easy for this material to come before court with no pre-emptive checks against tampering, giving the court full discretion on how it is considered? Or, do we want to have some questions that a court must ask when deciding whether to let such evidence come on record? In legal terms, the dilemma is about whether admissibility requirements for such electronic evidence should be relaxed or strict. 

In this post, I explain the evolution of law on this point, which shows that barring a brief decade between 2005 to 2014, the Supreme Court approved of strict admissibility requirements for electronic evidence. I argue that this approach was guided by a fear of the unknown when it came to technology, and ended up conflating issues of admissibility with those of a material's reliability. After 2014 though, courts have marked a silent retreat from this strict position, and in 2018 the Supreme Court suggested a return to the relaxed approach of 2005. The result of this back-and-forth is that trial courts across the country are uncertain about  what the law is, increasing the potential for manifest arbitrariness in how cases are processed. I offer my (utopian) take on how to solve this confusion, and conclude by urging for a speedy resolution of the present uncertainty to prevent costly failures of justice. 

Admissibility v. Reliability: The Judicial History of Electronic Evidence

Election cases have had a disproportionately large footprint in the judicial history of electronic evidence. In most of these, candidates had made recordings of their competitors violating election laws, and sought to use this recording as evidence in Court. How these cases were handled by the Supreme Court was the subject of a four-part series on the Blog, and I need not address that in much detail here. It suffices to say that while the Court often spoke about admissibility in those decisions, it paid lip-service to the statutory rules on the issue. Instead, the Court created new conditions beyond the text, which were concerned with the reliability of recordings and sought to exclude material that did not satisfy a minimum threshold of genuineness:

(1) The voice of the speaker must be duly identified by the maker of the record or by others who recognise his voice. In other words, it manifestly follows as a logical corollary that the first condition for the admissibility of such a statement is to identify the voice of the speaker. Where the voice has been denied by the maker it will require very strict proof to determine whether or not it was really the voice of the speaker. 

(2) The accuracy of the tape recorded statement has to be proved by the maker of the record by satisfactory evidence - direct or circumstantial. 

(3) Every possibility of tampering with or erasure of a part of a tape recorded statement must be ruled out otherwise it may render the said statement out of context and, therefore, inadmissible. 

(4) The statement must be relevant according to the rules of Evidence Act. 

(5) The recorded cassette must be carefully sealed and kept in safe or official custody. 

(6) The voice of the speaker should be clearly audible and not lost or distorted by other sounds or disturbances.

[Taken from Ram Singh v. Col. Ram Singh, 1985 (Supp) SCC 611]

A look at this test shows how the Court conflated issues of a material's admissibility with those of its reliability. Conditions on admissibility decide whether or not a court can look at material. Questions of reliability comes after the material is already before court, and affects how much weight can be placed on it. But since the decisions consistently referred to admissibility, it must be assumed that the idea was to create additional requirements to tighten statutory rules on admissibility, which were perceived as insufficient, with non-compliance warranting exclusion of the inadmissible material. 

The judiciary's multi-factor test arose at a time when Section 65-B did not exist in the Indian Evidence Act 1872. This provision, along with Section 65-A, was inserted in 2000, and provided a new procedure for admitting electronic evidence in court and address that judicial concern with a laxity on admissibility requirements in the statute. The new procedure did not differ much from the old — get originals if you can, but we will accept copies if that isn’t possible — but made some important tweaks. As Section 65-B(2) shows, the statute now carried specific checks at the stage of admissibility similar to the judicially created multi-factor test referred to above: 

(2) The conditions referred to in sub-section (1) in respect of a computer output shall be the following, namely :- 

(a) the computer output containing the information was produced by the computer during the period over which the computer was used regularly to store or process information for the purposes of any activities regularly carried on over that period by the person having lawful control over the use of the computer;

(b) during the said period, information of the kind contained in the electronic record or of the kind from which the information so contained is derived was regularly fed into the computer in the ordinary course of the said activities;

(c) throughout the material part of the said period, the computer was operating properly or, if not, then in respect of any period in which it was not operating properly or was out of operation during that part of the period, was not such as to affect the electronic record or the accuracy of its contents; and

(d) the information contained in the electronic record reproduces or is derived from such information fed into the computer in the ordinary course of the said activities.

..... 

Electronic Evidence: Anvar and the Strict Approach

Thus, the text of Section 65-B suggested that the legislature had introduced conditions in the statute that confirmed the judicially created tests that had emerged to regulate admissibility of electronic evidence. Crucially, though, when Section 65-B came before the Supreme Court for the first time in Navjot Sandhu [(2005) 11 SCC 600], it held that the provision had done no such thing. Rather, it held that Section 65-B was only one method for admission of evidence, which meant that the pre-emptive checks it provided were not mandatory. This was a clear indicator that the Supreme Court favoured a relaxed approach to allow relevant evidence and let the judge deal with it, rather than exclude it as inadmissible altogether. 

In 2014, the Supreme Court opted for a strict approach on accepting electronic evidence through a  decision by a Three Justices' Bench in P.V. Anvar v. P.K. Basheer [(2014) 10 SCC 473]. Anvar was also a case of alleged election malpractice, and the Court was asked to decide whether the material basis for the allegations — A/V recordings on a CD — was admissible or not. Despite these similarities with the old cases on tape-recorded evidence, Anvar did not simply apply those tests and decide the case. Instead, it correctly looked at the material from the prism of Section 65-B of the Evidence Act, and held that if parties were not getting originals then all copies must comply with the conditions of Section 65-B to be admissible. 

Bringing originals is not difficult with recordings, as all you need to do is get the original memory card etc. which was used in the recording device, along with the device itself. But with messages, emails, call records and the like, the originals are not in any person’s possession, but on the servers of corporations such as Google, Amazon or Airtel. Thus, for a vast chunk of electronic evidence, courts can only ever deal with copies. And in Anvar the Supreme Court held that to even look at any of this, Section 65-B had to be complied with, where compliance meant filing a contemporaneous certificate as provided under Section 65-B(4) (filed when the evidence comes to court). 

Anvar affirmed the strict approach on admissibility and overruled Navjot Sandhu in the process. In doing so, it harmonised the old cases on tape-recorded evidence with the new statutory provisions in Section 65-B. At the same time, it also prescribed a very harsh medicine for non-compliance with the law: excluding potentially relevant evidence from consideration. 

The Silent Retreat After Anvar

Anvar was flawed. Not because it held that Section 65-B is a mandatory requirement for accepting copies of electronic evidence, but because it held that a contemporary 65-B certificate is mandatory. This is plainly unsupported by the provision, which only provides for the certificate method as an option for convenience. 

Further, Anvar was also flawed insofar as the holding was allowed to apply retrospectively to pending cases and appeals, something that consequently placed all pending litigation in the country in a flux. If electronic evidence had been taken on record contrary to the single method that Anvar now prescribed, it had to go away. And since the method in Anvar made it impossible to cure defects — it needed a contemporary certificate — it meant many cases would fall to the floor. 

The initial chiseling of Anvar began with this perspective: courts read the judgment in ways to make the conditions more palatable so as to not imperil pending cases, but did not disagree with the view that Section 65-B was mandatory. A leading example of such an attempt was the Delhi High Court decision in Kundan Singh [Crl. Appeal 711/2014 decided on 24.11.2015], which enlarged the scope to cure defects by holding that Anvar did not require a contemporary certificate under Section 65-B to be filed. 

This chiseling continued with Sonu v. Haryana [(2017) 8 SCC 570]. Here, a Two Justices' Bench of the Supreme Court developed a different line to allow for curing the procedural defects that might result from Anvar. Focusing on the distinction between issues of "inherent admissibility" and "mode of proof", it held that non-compliance with Section 65-B was an issue that affected the latter. This required parties to raise this objection at the time of trial, and barred them from raising it in appeal. Effectively, it meant that the Court had shielded trial court verdicts from a retrospective application of Anvar, something it specifically flagged as a concern.

And in 2018, this chiseling fashioned a different artwork altogether when another Two Justices' Bench decided to take up many evidentiary issues in the case of Shafhi Mohammad [(2018) 2 SCC 801]. The case did not involve any questions on Section 65-B, but that did not stop the Court from "clarifying" the legal position on the same. It did so in two remarkable ways: first, it suggested that Sections 65-A and 65-B were not a complete code on the subject of electronic evidence, and second, it held that courts can relax the requirement of a certificate in the interests of justice as these are procedural provisions.

Working Towards Untangling the Knots

The chiseling, or wrecking, of the house built by Anvar in 2014 has been done either by High Courts, or by Supreme Court decisions passed by benches of lesser strength. This repeated clarification of the law by different courts is not an exercise in distilling the essence of a substance by repeated filtering. It is actually the opposite: a process which throws millions of pending into a violent churning, with the trial courts not knowing how to handle the material presented before them. 

How do we resolve this confusion over what shape should the law take? There are many answers to this. One of these is a clarification of Anvar to state that a contemporary certificate under Section 65-B is not required, and limiting the retrospective effect of that decision to save decided cases. This would address the most visible of sores. But it will not solve the problem itself. For that, we must repeal Section 65-B of the Evidence Act and bring in its place a new setup to untangle the nexus between admissibility and reliability that has contributed to the present state of affairs.

The judiciary's approach to focus on the accuracy of electronic evidence at admissibility itself worked at a time when such material was a rare oddity in trials. It was, as I have argued earlier, symptomatic of worries about misuse of unfamiliar technology that plague us all. Given the rarity of this material, as well its unfamiliarity, it makes sense that a system potentially excluding this evidence at the outset was not seen as problematic. Section 65-B reflects this reality — the product of a legislation that went through its gestation in the 1990s. However, the landscape has changed so dramatically in the last two decades that both aspects of rarity and unfamiliarity in relation to electronic evidence have almost completely disappeared.

This transformation makes a legal system which possibly excludes this material as evidence appear quite unfavourable and out of sync with common sense, and explains the retreat from Anvar. It also warrants that the system give up this approach which considers issues of accuracy and reliability at the stage of admissibility itself and ousts relevant material. This ultimately increases the already existing gap between judicial verdicts and the truth as parties might perceive it. Instead, it makes sense to revert back to the standard method for accepting evidence, found in Sections 60 to 65 of the Evidence Act. In this model, admissibility is concerned about the provenance of the material, with a preference for getting originals with the law allowing copies if the circumstances require. But the concerns over accuracy and reliability, which are at the heart of Section 65-B(2), would still be left for courts to consider when evidence is being appreciated and arguments being heard. 

Conclusion
The Indian Parliament took great inspiration from Section 5 of the U.K. Civil Evidence Act of 1968 while drafting Section 65-B: both provisions were almost identical. I use the past tense because the U.K. provision has since been repealed, with that jurisdiction adopting a relaxed approach on the point of admissibility. Here, I suggest that India should follow suit, and consign Section 65-B of the Evidence to the realm of history.

The chances of that are not very high, which is why the impetus then shifts to the judiciary bringing about changes to Anvar for making it more palatable — a petition seeking clarifications is presently pending. Even so, I would suggest that a return to the 2005 position in Navjot Sandhu should also be considered, which suggested that Section 65-B cannot be treated as the only method for accepting electronic evidence. 

Whatever way the cards fall, clarity is urgently needed. I exaggerate, but the multiplicity of judicial opinions from high above in the wake of Anvar has perpetuated a situation where nobody on the ground knows what the law is with any certainty. Rather than more episodes such as Shafhi Mohd., a decision which changes the law on electronic evidence in a case which had nothing to do with it, what we need is a firm resolution to prevent the arbitrariness from causing more failures of justice.  

Electronic Evidence - Part 3

The treatment of electronic evidence in India has been the subject on this blog for the past few posts. This has been done by chronologically discussing the cases involving questions on the admissibility of tape-recorded evidence. From having no tests in the beginning, the Supreme Court by the 1980s had begun to impose strict admissibility requirements for such evidence. This was, in my opinion, largely driven by the reliance judges placed on the English decision in R v. Maqsud Ali. Frequent invocation of the requirements imposed by the Court of Appeals in that case to judge electronic evidence lend credibility to the view that the same tests now applied to India. This was put beyond doubt by the Court in Ram Singh v. Col. Ram Singh [1985 (Supp) SCC 611].

Ram Singh and Tying Loose Ends

Another election dispute, another judgment on tape-recorded evidence. But here, all three judges gave separate opinions in what was a 2-1 split. The majority formed by Fazal Ali and Mukharji JJ., held the tape-recorded statements were inadmissible, whereas Vardarajan J. disagreed. 

Right at the outset of his discussing the evidence, Fazal Ali J. observes: 

Thus, so far as this Court is concerned the conditions for admissibility of a tape recorded statement may be stated as follows:

(1) The voice of the speaker must be duly identified by the maker of the record or by others who recognise his voice. In other words, it manifestly follows as a logical corollary that the first condition for the admissibility of such a statement is to identify the voice of the speaker. Where the voice has been denied by the maker it will require very strick proof to determine whether or not it was really the voice of the speaker.

(2) The accuracy of the tape recorded statement has to be proved by the maker of the record by satisfactory evidence - direct or circumstantial. 

(3) Every possibility of tampering with or erasure of a part of a tape recorded statement must be ruled out otherwise it may render the said statement out of con text and, therefore, inadmissible.

(4) The statement must be relevant according to the rules of Evidence Act

(5) The recorded cassette must be carefully sealed and kept in safe or official custody. 

(6) The voice of the speaker should be clearly audible and not lost or distorted by other sounds or disturbances.

This observation notably expanded on the course plotted earlier. To lend greater authority to his observation, the judge referred (again) to Maqsud Ali, and also to American opinion on the issue. He discarded the evidence as, although relevant, it failed the other conditions of reliability in his six-fold test. Interestingly, Mukharji, J. in his separate, concurring opinion discussed Maqsud Ali to observe that the court there held "there cannot, however, be any question of laying down any exhaustive set of rules by which the admissibility of such evidence should be judged." The dissent did not provide a rival six-fold test, but merely stated that "tape-recorded evidence is admissible provided that the originality and the authenticity of the tape are free from doubt." The dissent, thus, simply found the facts sufficient to establish all that the majority also wanted to admit the electronic evidence.

Infidelity to the Text

The six-fold test conceived by Fazal Ali, J. appeared an easy device for courts to employ in dealing with electronic evidence at large. The three judges in Ram Singh extensively considered arguments on admissibility of Secondary Evidence, i.e. tape-recorded evidence in this case. Yet, the three separate judgments are astonishing for their failure to even mention Section 65 of the Evidence Act, the basic provision governing the admissibility of Secondary Evidence. In fact, there is no discussion about the nature of the evidence at all. Rather than turn to the text of the governing law, the Supreme Court repeatedly referred to, and relied upon, authorities from the UK and USA. 

The importation of foreign tests to the Indian legal system has been criticised often. There are unique challenges to each legal system and legislators and courts alike have been all too keen to ignore this and adopt 'best practices' to fill the gaps. The same issue arises here: the UK and India fundamentally differed in their evidence laws as the former had no statute! The Court of Appeal in Maqsud Ali was merely following common law traditions of developing law in the absence of statute. This did not apply to India because the Evidence Act squarely covered the field. There were no gaps here - having characterised the evidence as either Primary or Secondary, the Court was obligated to turn to the statute. If it considered statutory requirements too relaxed, it could recommend amendments (which finally happened in 2000). The admissibility-heavy criteria in Maqsud Ali might have appeared necessary because of the jury-trial in the UK. The judge ensured only the best and most accurate evidence reached the jury, which was solely qualified to determine its reliability. But where the judge himself determines the reliability, such entry-barriers don't have such a beneficial effect.

Conclusion

Surprisingly, the Ram Singh tests continue to be used by courts despite the amendment of the Evidence Act to specifically deal with electronic evidence [see, e.g., Tukaram Dighole v. Manikrao Shivaji Kokate (2010) 4 SCC 329]. This supports the argument I am advancing here: that the Supreme Court has been extremely unfaithful to the text of the statute, only to address its fears over the genuineness of electronic evidence. The most recent example of this, Anvar P.V. v. P.K. Basheer [(2014) 10 SCC 473], will be the focus of the last post in this series.

Electronic Evidence - Part 2

In the last post, I put across the idea that the Supreme Court has had a very big role to play in shaping how electronic evidence is treated in India. This influence, I argue, has in some instances led to departures from the statute to reinforce the Court's opinion of the proper state of affairs. Fears about the reliability of such evidence are seemingly underpinned by a desire to ensure the best evidence makes its way to court. But an overemphasis on these issues has led to disturbing a seemingly delicate balance that the Evidence Act maintains between admissibility of evidence and determining its reliability. Here, I develop this thought by showing the progression of law regarding, the admissibility of tape recorded statements. 

Before proceeding further, a fun fact about the cases to be discussed. These were either election disputes, or sting operations, where tape recorded statements were adduced in evidence against the accused. None of them were instances where the evidence was relied upon by the accused himself to put forward his defence. Given the (now declining) importance paid to the presumption of innocence in our criminal justice system, one wonders whether gradual introduction of increasingly strict admissibility rules might have been driven by this commonality across these seminal cases. 

Partap Singh and the Beginning

The first notable case was Partap Singh v. State of Punjab [(1964) 4 SCR 733]. The Appellant, a Civil Surgeon, had been found guilty of unlawfully accepting 16 Rupees and been placed under suspension. The High Court held against him in Article 226 proceedings, and the Appellant thus approached the Supreme Court. A 3-2 split held in favour of the Appellant. Now, the Appellant had heavily relied upon tape recorded statements to support his case, which the High Court refused to admit in evidence owing to fears of tampering. This approach was rejected by all five judges in the Supreme Court. Tellingly, the Majority observed that "There are few documents and possibly no piece of evidence which could not be tampered with, but that would certainly not be a ground on which Courts could reject evidence as inadmissible or refuse to consider it." Nowhere in the judgment does the Court lay down tests to consider the admissibility of these records. The Majority simply observed that: "Doubtless, if in any particular case there is a well-grounded suspicion, not even say proof, that a tape-recording has been tampered with, that would be a good ground for the Court to discount wholly its evidentiary value." Thereafter, it is stated that the State in this case had applied to the Court to create copies of the original records supplied by the Appellant to satisfy itself on the genuineness of the record. Eventually, the State did not deny the genuineness of the recordings.

Although Partap Singh did not lay down a test, there is a passage that seems to be the basis of the later forays of the Court. While noting that the State made its copies of the tapes, the Majority observed that this was done "to verify (a) whether the voice recorded was that of the person whose voice it professed to be; (b) whether there had been any interpolations or omissions; and (c) whether there had been any other tampering with the records."  A reading of the entire judgment leaves no doubt that this is not the Court laying down a test. It appears nothing more than a mere description of what the State would have done. Nonetheless, this position soon came to represent what the Court required to be done in such cases.

'Salim Pan Lao' and the Malkani Phone-Tapping
The concept of entrapping accused persons took a whole new leap with tape-recording technology. It was not long before someone challenged its use, and the Supreme Court specifically addressed this issue in Yusufalli Esmail Nagree v. State of Maharashtra [(1967) 3 SCR 720]. The Appellant had arranged to meet a public servant (here, a clerk) at his residence for paying a bribe. The clerk had already informed the Police, and so a trap was laid. When the Appellant produced the money to hand it to the clerk, he gave that signal 'Salim Pan Lao', and the Police came rushing into the room to catch the Appellant red-handed. The tape recordings were used as evidence to convict the Appellant, who thus took his challenge to their admissibility to the Supreme Court. The three judges cited Partap Singh along with R v. Maqsud Ali [(1965) 2 All ER 464] and held:

"If a statement is relevant, an accurate tape record of the statement is also relevant and admissible. The time and place and accuracy of the recording must be proved by a competent witness and the voices must be properly identified. One of the features of magnetic tape recording is the ability to erase and re-use the recording medium. Because of this facility of erasure and re-use, the evidence must be received with caution. The Court must be satisfied beyond reasonable doubt that the recording has not been tampered with."

The tampering fear is laid bare here. Because of this, from requiring no tests in Partap Singh, the judges in Yusufalli held that the recording's (i) time, (ii) place, and (iii) accuracy must be proved, and voices therein properly identified. Given that this was a criminal case, the Court clarified that all the above must be proved beyond reasonable doubt. Still, the judgment is unclear on whether these observations concerned the admissibility of tape recordings, or merely indicated factors that would affect their reliability. An equally important development here was the reliance placed on the English decision of Maqsud Ali, something I will return to later.

V.V. Giri and the Presidential Election
Apart from trap cases, tape recordings were quite common evidence in election disputes. In N. Rama Reddy v. V.V. Giri [1970 (2) SCC 340], the Court was concerned not purely with the admissibility question, but whether tape recorded statements could be admitted during cross-examination of witnesses to contradict them. The case was sensational as it concerned the Presidential Elections for the country. The Petitioner had sought to confront witnesses using their earlier tape recorded statements and so the Court was required to interpret Sections 146 and 155 of the Evidence Act. After observing the decisions of the U.K. and India, the Constitution Bench here found no difficulty in concluding that such evidence was admissible even for purposes of cross-examining the witness. For reasons best known to the judges, they went no further in discussing the admissibility requirement than by citing various decisions of the High Courts and Supreme Court of India, along with other decisions from English Courts. It remains unclear, therefore, whether the Constitution Bench favoured the view in Partap Singh or that of Yusufalli.

Consolidating a Test: Malkani and Ziyauddin 
In R.M. Malkani v. State of Maharashtra [1973 (1) SCC 471], the Appellant (a Coroner) was demanding a bribe from a Doctor to give a favourable report. There were several conversations over the phone between the Appellant and witnesses conveying a demand and on how the bribe was to be paid. When the witnesses approached the Police, the telephone lines of a witness were tapped to record these incriminating calls between him and the Appellant. While the bribe was ultimately not paid, the Appellant was charged with an attempt to demand and accept illegal gratification. The Two judges in Malkani cited the previous decisions on the point mentioned above without engaging with their reasoning. No provision of the Evidence Act was mentioned either. Rather, at paragraph 23 of the SCC reported version we find the Court stating tests that tape recorded evidence must satisfy to be admissible in evidence: (i) the voices of speakers need identification, (ii) accuracy of the record is proved "by eliminating the possibility of erasing the tape record". The near-parity with Maqsud and Yusufalli is apparent. But now, the judges went further and clarified that these tests determined the admissibility of evidence, a question perhaps left open afterYusufalli.

Five judges in Ziyauddin Burhanuddin Bukhari v. Brijmohan Ramdass Mehra [(1976) 2 SCC 17] were required to decide an election dispute which again involved reliance on tape recorded evidence. The High Court had admitted the evidence after having tested that (i) the voices of speakers were identified, (ii) there was no erasure and (iii) the conversation was relevant to the dispute. In upholding the High Court's decision to admit the evidence, the Constitution Bench stated that "these requirements were deduced by the High Court from R v. Maqsud Ali ...". Interestingly, the Court did not cite Malkani although the requirements of Maqsud had already been incorporated in India by that decision. Nor did the Bench provide provide any update/clarification on the tests laid down; they seemed content with status quo.

Conclusion
Between 1965 and 1975, three Constitution Bench decisions considered the admissibility of tape recorded statements, and said very different things. The gradual development of strict admissibility criteria for electronic evidence is clear, and it certainly seems driven by the fear of tampering this evidence. An underlying commonality missed out by commentators, however, is the heavy reliance on the English Court of Appeal decision in Maqsud. The Supreme Court imported those tests over the course of two decades. Next time, I try and argue how this was a mistake. I also discuss how the Court compounded this mistake: the decision in Ram Singh v. Col. Ram Singh [1985 (Supp) SCC 611].