When Digital Evidence Can Be Fake: The Challenge of Proving Authenticity in AI-Generated Evidence : Author: CHITHRA C
The rapid development of AI increased the possibility to create and manipulate realistic digital content and that increased the challenge to prove the authenticity of the digital evidence. This article examines the legal validity of digital evidence, with relevant acts and judicial decisions. It focuses on the challenges posed by AI and also highlights the need for stronger authentication practice s to prevent fabricated digital evidence.
ARTICLE
Abstract
The rapid development of AI increased the possibility to create and manipulate realistic digital content and that increased the challenge to prove the authenticity of the digital evidence. This article examines the legal validity of digital evidence, with relevant acts and judicial decisions. It focuses on the challenges posed by AI and also highlights the need for stronger authentication practice s to prevent fabricated digital evidence.
Key words: AI, digital evidence, BSA, IT Act, Chain of custody, metadata
Introduction
The widespread use of the internet, cloud computing, artificial intelligence, blockchain technology cryptocurrencies and digital payment systems has significantly contributed to economic growth and technological innovation. However, alongside these developments have led to emergence of new opportunities for criminal activities and the manipulation of digital evidence which is used in crime investigations. India has enacted significant legislation, including the IT Act, BNS, BSA and DPDP Act, the rapid revolution of emerging technologies continues to create legal and practical challenges. One such challenge is the authenticity of digital evidence, particularly AI generated and manipulated images, videos and audio. This raises an important question regarding the Genuity of digital evidence.
Electronic evidence & legal validity
Under the BSA, 2023 recognize electronic record as documentary evidence. It includes any information generated, stored, received or transmitted through any electronic means. The BSA gives electronic record the same legal status as physical document under the section 61, 62 and 63 of the acts. Under the IT act 2000, in section 2(t) of the act given the definition for the electronic record, as means any information that exist in digital form and can be retrieved from a computer, mobile phone, served or other electronic device is an electronic record.
The admissibility of the electronic evidence record allowed to be received as evidence in court, the proof of the credibility of the evidence must be proved in court, so under section 63 only allow the admissibility of the evidence not the proof value of the evidence. Under the 63(4) of the acts the submission of the certificate from the authoritative person is mandatory for to consider the record as evidence if there is original electronic record given then the necessity of the certificate is invalid. In state (NCT of Delhi) v. Navjot Sadhu[1] honourable supreme court held that electronic evidence could be admitted without section 65B certificate by relying on general ruled of secondary evidence. In the famous case Anvar v. Basheer 2014[2] in this supreme court overruled the Navjit Sandhu and held that section 65A and 65B under the Indian evidence act formed a complete code for electronic evidence for the secondary electronics the certificate is mandatory. In the recent judgement by the supreme court under the case Pune Bar association v. UOI 2026[3] court upheld the validity of section 63(4
Understanding AI-Generated & Manipulated Evidence
Digital evidence has become important for modern litigation and criminal investigation. CCTV, mobile phone data, WhatsApp messages, emails, photographs, audio recordings and social media posts can be used to help establish facts before the courts. However, the development of generative AI has changed the nature of evidence. AI can now create realistic photographs, videos and voices of people who's never appeared in the recorded situation. Deepfake technology can also alter genuine content in a way that may be difficult to identify merely by looking at or listening to it. The problem is not simply whether digital evidence exists or not; it's the validity of the evidence. This creates a serious challenge for courts because digital evidence can be copied, edited, manipulated, and now even completely generated through AI.
Why AI makes Authenticity more difficult
Traditional digital evidence begins real time events, where the cctv footage collected and produced before the court, but now days the influence of generative AI can manipulate the real time footage also. That where the chain of custody becomes important, the court needs to know where the evidence comes from, who collected, what device originally contained it, whether it was altered and how it reached the court. Metadata is usually used to find the source of the data, but also the production of metadata can't solely prove the originality of the evidence.
Another difficulty is the rapid development of AI itself. Detection software can identify the manipulated content, but these tools have limitations. The detection result may therefore need to be supported by forensic examination.
Role of court
In Tomaso Bruno v. State of Uttar Pradesh, the supreme court held that the scientific tamper and technological methods should form, part of modern investigation and recognized the importance of electronic evidence including cctv the concept of the secondary evidence to best evidence. This principle plays a significant role in case of AI related matters, and court may need digital forensic evidence to examine the originality of the evidence. Expert analysis can help to determine the value of the evidence. But forensic technology should not replace judicial scrutiny. The court must consider the technical finding together the circumstances of the case and other evidence.
The way forward
The increasing use of AI doesn't make digital evidence suspicious in every case. The process of authentication must be stronger to preserve the original devices and files, whatever possible, to maintain a clear chain of custody. Courts should also have access to competent digital forensic experts to check the authenticity of the evidence.
There is also needed clearer guidelines to identify the deepfakes and AI generated evidence. Lawyers, judges, and investigating officers require greater technical awareness to understand the limitation of digital evidence for modern legal practice.
Conclusion
AI has made digital evidence both more useful and more difficult to verify. The existing legal framework provides important safeguard for electronic records under section 63 of BSA,2023. However, AI-generated evidence raises a different level of concern about the originality of the evidence. The court not only checks the digital evidence but also the source, integrity, and connection with real events related to the case. As technology continues to develop, the legal system must strengthen authentication and forensic practice to identify genuine digital evidence to avoid the fabricated evidence.
[1] State (NCT of Delhi) v Navjot Sandhu (2005) 11 SCC 600
[2] Anvar PV v PK Basheer (2014) 10 SCC 473
[3] Pune Bar Association v Union of India WP (C) No 599/2026