By- Aditya Kumar Singh
The Indian Evidence Act, 1872 has a strict distinction between fact and evidence, and for any document, electronic record, or oral testimony to become evidence and be considered during the final adjudication, it has to first satisfy three pre-conditions in evidence law, which can be formulated and structured as three doors of evidence. First, the relevance of the fact sought to be proved, second, the admissibility of the evidence to be used to prove the fact, and third, the reliability of such evidence, which the Evidence act leaves to the judge’s discretion. There has been an ongoing debate on the legality of sting operations and their subsequent admissibility as evidence in a court of law, with different courts taking different views. But, beyond this second door of admissibility lies reliability, an issue that becomes exponentially more important in cases where the courts are supposed to be guided by this audio or video recording obtained without the knowledge of the person who was the subject of such recording. Further, an even more nuanced aspect of such a recording is the statement that it carries, and whether such statement is classified as an admission or a confession by the court. This article seeks to highlight the manner in which the Indian courts have been mis-handling admissions made in audio-video recordings procured through sting operations, and also recommend that reliability be made a mandatory pre-condition for even considering admissibility of such evidence.
Triangulating Admissions
The problem of hearsay arises when a third person tries to testify to a statement made by a person not before the court. Although audio-video recordings of such statements give better accounts of the utterances than any witness ever could, but there still exists a problem common to hearsay and sting operations. Professor Tribe in his article on hearsay has aptly captured this as the problem of sincerity and ambiguity. While Professor Tribe has elaborately explained these problems in the context of hearsay, a leading example of the problem of sincerity in sting operations is the sting conducted on Ajay Kumar Katara, the star witness for the prosecution in the Jessica Lal murder case, who saw the deceased in the company of the accused just before her death. In the sting operation the witness had admitted to giving a false testimony against the accused at the behest of his political rivals, he admitted to cooking up the entire story, and affirmed that the motive was to bring disrepute to the accused’s father. On being questioned about the same in court, the witness stated that the admissions in the sting were false. He justified the same by saying that the accused’s family had been threatening his family, and the statements made in the sting were to merely mislead the accused’s family into believing that he was amenable to changing his testimony. Though questions were raised about the authenticity of the sting, neither did the court allow cross-examinations of the parties involved in the sting operation nor did it get it authenticated by a forensic laboratory. The court in the exercise of its discretion relied on the justification tendered by the witness, opining that since the witness had never changed his story in cross-examination and there have been threats against his family earlier as well, his justification seemed reasonable.
An example of the second problem, ambiguity, is the landmark case of R.K. Anand v Registrar, Delhi High Court, where the issue of whether or not the public prosecutor colluded with the defense involved the interpretation of the phrase ‘bade saheb’. The public prosecutor contended that he used the phrase to refer to a senior police official, while the news agency and Mr. Kulkarni, who conducted the sting operation contended that it was used to refer to the defense counsel, R.K. Anand. The court, once again, did not feel the need to have the recording authenticated by a forensic laboratory even though specific requests were made to this effect, and went on to decide the interpretation on the basis of another unverified audio recording. But a contrasting approach taken by the Allahabad High Court in the case of Raj Veer Singh v State of U.P., showcases the correct method of handling sting operations as evidence. A news agency had conducted a sting operation on the officials of the Forensic Laboratory, Agra, alleging that the right price could buy a favorable report.The alleged sting operation had an ambiguity with respect to the statement of the subject where he used the words “two plus”, the news agency depicted this in a manner which made it seem like the conversation pertained to asking for more than two lakhs as bribe money. But the subject contended that the words used were in the context of certain tests conducted for determining the level of sugar and albumin in the urine. The court observed that the issue of authenticity of the recording was central to the case and needed to be considered before reaching any substantive conclusion. They relied on a report from the Anti-Corruption Unit’s forensic experts that unequivocally declared that the recording was edited and morphed, with multiple scenes being omitted from even the supposedly unedited raw footage. Thus, the court dismissed the allegations of bribery but still levied a minor punishment on the laboratory officer for even entertaining people who had the intention of bribing him.
The Jessica Lal case is one of the most controversial murder trials in the history of this country, and the case of R.K. Anand is still the lex loci in matters of sting operations as evidence and their admissibility. Both these cases are examples of the amount of discretion the court enjoys in determining the veracity and context of an admission. But, both these cases are also examples of the reckless manner in which the courts handle audio-video evidence, to the extent where the authenticity of such electronic evidence is being ascertained on the basis of how it looked to the naked eye. Even though specific contentions were raised against the authenticity of these recordings and requests were made to have them checked by a forensic laboratory, the court did not deem it necessary to have them verified. On the other hand, the case of Raj Veer Singh v State of UPis a testament to the fact that there is merit in having recordings authenticated before taking them into consideration, even if the judge is extremely tempted by its ostentatious probative value. The next part of the article shall deal with the far-reaching impact of the aforementioned recklessness of the court in the 21st century.
Standard for Audio-Video Evidence
The Supreme Court in the case of Ram Singh v Col. Ram Singhlaid down the standard for admissibility of tape recordings, where the most important one was verifying the identity of the speaker by comparing his voice sample with the voice in the recording, and ensuring that all possibilities of tampering have been ruled out. This standard is utilized by the apex court to this date, even in cases of sting operations involving an audio-video recording. But with the advent of technology and introduction of Deepfakes as evidence in court proceedings, this archaic modus operandi is bound to be defeated. This is because, in addition to super-imposing the face of the subject on another person to get the requisite facial expressions and lip movement, Deepfakes also involve isolating words from former speeches and stitching them together to form complete sentences in the voice of the subject. An example of the same is the fake Barack Obama speech prepared by the researchers at the University of Washington.
Further, even if the judiciary were to strictly comply with the current legal requirements for ascertaining authenticity it still might not be enough. This is because the authentication process involves looking for minute digital alterations like frame dropping or duplication in addition to voice recognition. Since the technology available now can create entire statements in someone else’s voice without the need to cut or edit the actual recording, this safeguard of checking for digital footprints and voice recognition will also be less effective. Admittedly, there still exist some methods for checking whether a recording is genuine or a Deepfake. But in cases of videos, the electronic file is already greatly compressed, further, in cases of sting operations since the recording is through a spy-camera, the resolution of such a recording is relatively lower which makes retracing the digital footprints and alteration in the size of the RAW footage even more difficult.
Currently, the legal requirement for audio-video recordings does not mandate authentication from a forensic expert. The requirement for all electronic evidence, ranging from electronic mails to call detail records to even audio-video recordings is just a 65-B certificate. But, the authentication provided by such a certificate should not be enough in cases of sting operations for two major reasons, first, the person submitting the certificate is only supposed to state that the recording is true to the best of her/his knowledge.In cases where a doctored audio-video recording is being submitted as evidence before the court, the person is already committing an act punishable under the IPC by fabricating evidence. Therefore, submitting a 65-B certificate to increase the recording’s credibility isn’t difficult. Second, on the scale of probative value v/s prejudicial value, audio-video recordings is more prejudicial than any forged electronic mail or call detail records. This is because visual representation has a greater ability to cause prejudice to the defendant’s case. As can be seen from the case of Raj Veer Singh v State of U.P., where even though the defendant did not commit the crime of accepting a bribe but just because the video footage showed him engaging with the people who were trying to bribe him, the court felt that some form of punishment was still necessary. The courts already enjoy a lot of discretion in determining the meaning and context of admissions made in the recordings. But the technological authenticity of a recording is a scientific fact, capable of being ascertained by the right expert. Thus, the courts using their discretion to determine such authenticity is not just reckless, but an opportunity for the litigants to game the criminal justice system.
Although professor Tribe in his article offers delayed cross-examination of the declarant as a solution to the problem of sincerity and ambiguity in cases of hearsay, but this solution cannot be applied to admissions made in sting operations for two major reasons. First, as highlighted above, the current standards of the Indian criminal justice system leave room for the possibility of introducing completely fabricated recordings as evidence. Something that has also been done in the past where a school teacher was wrongly accused of running a prostitution racket on the basis of a staged sting operation. Second, since the standard for admissibility of admissions is already lower than that of confessions, and the apex court has held admissions to be substantive evidence, introduction of such unverified recordings into the court proceedings will unfairly prejudice the case of the defense. This is because, on the face of it such recording might seem highly probative, tempting the judge to admit the same. But the fact remains that its admissibility and probative value are all dependent upon its reliability and authentication. Thus, before burdening the defense with the task of negating such an incriminating piece of electronic evidence, the courts should mandatorily verify its authenticity.
Conclusion
The current state of affairs is not only insufficient to responsibly handle and consider evidence procured through sting operations, but also dangerously vulnerable to being misled by completely fabricated evidence. Evidence collected through sting operations is incriminating enough to place the defense at a strategic disadvantage from the very start of the trial. Thus, in cases where the admission of such incriminating facts is to be made admissible in the court proceedings, the courts should tweak the doors of evidence a bit to make reliability a mandatory pre-condition for even considering the issue of admissibility of such evidence.
[The author is a third-year student at NLS, Bangalore]