#FactCheck - AI-Generated Image of Abhishek Bachchan and Aishwarya Rai Falsely Linked to Kedarnath Visit
A photo featuring Bollywood actor Abhishek Bachchan and actress Aishwarya Rai is being widely shared on social media. In the image, the Kedarnath Temple is clearly visible in the background. Users are claiming that the couple recently visited the Kedarnath shrine for darshan.
Cyber Peace Foundation’s research found the viral claim to be false. Our research revealed that the image of Abhishek Bachchan and Aishwarya Rai is not real, but AI-generated, and is being misleadingly shared as a genuine photograph.
Claim
On January 14, 2026, a user on X (formerly Twitter) shared the viral image with a caption suggesting that all rumours had ended and that the couple had restarted their life together. The post further claimed that both actors were seen smiling after a long time, implying that the image was taken during their visit to Kedarnath Temple.
The post has since been widely circulated on social media platforms
Fact Check:
To verify the claim, we first conducted a keyword search on Google related to Abhishek Bachchan, Aishwarya Rai, and a Kedarnath visit. However, we did not find any credible media reports confirming such a visit.
On closely examining the viral image, several visual inconsistencies raised suspicion about it being artificially generated. To confirm this, we scanned the image using the AI detection tool Sightengine. According to the tool’s analysis, the image was found to be 84 percent AI-generated.
Additionally, we scanned the same image using another AI detection tool, HIVE Moderation. The results showed an even stronger indication, classifying the image as 99 percent AI-generated.
Conclusion
Our research confirms that the viral image showing Abhishek Bachchan and Aishwarya Rai at Kedarnath Temple is not authentic. The picture is AI-generated and is being falsely shared on social media to mislead users.
Related Blogs
Introduction
On June 11, 2026, the Ministry of Home Affairs (MHA) India released one of the most critical Indian government advisories concerning cybersecurity by the Indian Cyber Crime Coordination Centre (I4C) under the National Cybercrime Threat Analytics Unit (NCTAU) concerning the immediate and escalating threat posed by the weaponization of generative artificial intelligence to forge synthetic biometric identities capable of bypassing the existing facial verification mechanisms in India. This advisory is arguably one of the most explicit Indian government recognitions of the deep-seated threats associated with AI-generated deepfakes in the country’s digital financial infrastructure. As many Indian financial service providers embrace facial recognition and biometric verification systems for customer onboarding and authentications, the myth that biometric traits are in themselves secure is slowly unraveling.
The advisory states that cybercriminals are deploying sophisticated AI tools to forge such credible digital simulacrums that exhibit such a precise similarity of facial expressions, eye movements, eye blinks, head movements, and voice patterns that they are virtually indistinguishable from the originals for identity verification mechanisms. Such a confluence of easy AI technology, mass onboarding of digital identities, and underdeveloped infrastructure to detect these synthetics requires urgent regulatory, institutional, and technological intervention.
The I4C Advisory: Core Findings and Threat Architecture
In its advisory, NCTAU describes a complex, multi-step attack chain used by scammers to capture biometric information and perpetrate fraud using everyday social interactions. The attackers typically use social media accounts, chat messengers, online job applications, dating applications, or direct phone calls to reach their targets. These interactions are presented as innocuous, such as for video calls, job interviews, identity checks, or just normal conversation with the intention of recording facial and vocal data.
During these interactions, victims may be asked to perform gestures commonly seen in legitimate video calls, such as look directly at the camera, blink, turn their head, or say specific phrases. However, the perpetrators record this video feed without the victim's knowledge and then use deep learning generative AI technologies to process it. Through methods such as Generative Adversarial Networks (GANs) and diffusion models, the scammers create photorealistic synthetic duplicates of the target, capable of mirroring all physical and vocal attributes, such as facial expressions, blinking patterns, head movements, and even voice tones.
The advisory explicitly states that these synthetic identities can be used for a variety of fraudulent activities, such as spoofing face authentication systems, circumventing liveness detection checks, successfully completing video KYC, enabling fraudulent account recovery processes, and illegally accessing bank and financial services. NCTAU also cautions that these voice deepfakes may be paired with facial deepfakes in an attempt to undermine multi-modal authentication methods, and the occurrence of related SIM-swap attacks can eliminate the last layer of security in OTP verification and facilitate a complete account compromise.
The scale of India's Digital Financial Ecosystem
The scale of I4C's detected threat can be better understood by considering India's entire digital financial landscape. In 2025 India has witnessed over 228 billion UPI transactions, with 21.63 billion in December alone, an annual growth rate of 29% from 2024, and an active user base of over 500 million by the beginning of 2026. Furthermore, total e-KYC transactions by April 2025 have exceeded 2,393 crore, and thus, it can be seen the extent to which these aspects of finance (banking, insurance, and credit) are now conducted via remote digital verification. The transformation, although instrumental in increasing financial inclusion, has, according to some analysts, created an attack surface of historic scale. As hundreds of millions more become financially integrated via the very same channels that now form the country's infrastructure and systems of identity, the threat from identity-based fraud becomes astronomically large.
Indian government data further illustrates the extent to which such frauds are a growing concern. Cybercrime cases jumped 42% year-on-year to 2.27 million in 2024, resulting in losses amounting to nearly 228.45 billion. Within that, 1.34 million UPI cases, worth 1,087 crore, occurred in FY2024 alone, while cybercrimes in general soared from 260,000 cases in 2021 to nearly 2.8 million by 2025, totaling cybercrime losses of 22,931 crore.
How Do Deepfakes Defeat Biometric Systems?
Deepfake fraud, in particular, is extremely difficult to counteract due to the direct attack it poses on the assumptions underlying traditional verification systems. Passive techniques for verifying a live person from a static photo or video existed that primarily looked for similarities in textures, lighting, and geometrical properties or challenged subjects to perform an action in real-time. But the generation of real-time face swapping that contains blinks, head motion, and speaking can now be produced on even cheap machines. Cybercriminals can exploit these by using virtual camera drivers to "inject" the false image feed into the live verification session, nullifying any passive liveness checks. Data from the industry clearly shows the extent of this problem: iProov, a leading authenticator, documented a 7.8-fold rise in injection attacks in 2024; Jumio noted an 88% increase in deepfake-induced fraud in 2025; and voice-deepfake attacks on financial call centres saw a 6.8-fold increase in 2024.
Gartner had also predicted that 30% of organizations would have lost trust in facial verification alone by 2026, and work by Kubam (2024) confirmed a lack of multi-factor authentication such as cross-validation of biometric, document, and device integrity signals used within KYC platforms. Such fears have been corroborated by FATF's 2025 Horizon Scan, which classified deepfakes as an emerging threat to the AML/CDD framework and digital identity verification.
Recommendations by I4C
I4C's advisory goes beyond merely warning about threats and lists actionable recommendations to both institutions and citizens. Banks, NBFCs, fintech companies, and onboarding platforms have been advised to incorporate advanced deepfake and synthetic content detection techniques into their verification flows, given that first-generation liveness checks are not enough. They should employ a multi-modal strategy that considers face features along with the device, network signals, behavioral biometrics, and alignment of face and voice. They also have been advised to make a more robust upgrade of their onboarding and verification platforms, as much of the current remote verification architecture was built in a less sophisticated threat context. This aligns with the KYC Master Direction of the RBI that specifies end-to-end encryption, IP-based access controls, geotagging, and technology platforms and systems are to be upgraded frequently. Citizens are advised by I4C to keep their biometric information secure; be careful of unsolicited video calls and online interviews; keep an eye on transaction-related SMS and emails; and report suspicious instances through the National Cybercrime Reporting Portal and through the telephone number 1930. It is clarified that this advisory aims to create awareness of developing AI-based identity fraud schemes, and it is not a declaration that any specific organization, platform, or service is vulnerable.
The Legislative Dimension: India's Evolving Response to Synthetic Media
The problem highlighted by I4C is evolving in a heavily legislated environment, not a legal void. The first-ever legal definition of "synthetic media" in India came into force in the Information Technology Amendment Rules 2026 on February 20, 2026. These rules oblige significant platforms to remove deepfakes and non-consensual intimate media within three hours and two hours, respectively, or lose their safe harbor protection under Section 79 of the IT Act. While the provision focuses on harm stemming from content, this creates a new legal and normative precedent on dealing with AI-induced deception. However, financial frauds facilitated through deepfakes are not content but involve the use of remote identity verification and customer onboarding systems, which require specific technical standards. The overall policy environment when viewed in light of the FATF Horizon Scan, RBI KYC rules, and recent I4C advisory already offers significant scope to define and introduce mandatory deepfake detection and identity assurance standards even before these are explicitly legislated.
Institutional and Technical Recommendations
- For Financial Institutions and Fintech platforms: The existing verification systems (liveness detection) must be replaced with multi-layered deep-fake detection processes, including injection attack detection, behavioral biometrics, cross-modal facial and voice verification, device integrity check, and hardware attestation during onboarding itself.
- For Regulators: The RBI and Ministry of Home Affairs should work together to release technical standards that specify minimum deepfake-detection requirements for video-KYC and remote onboarding systems in line with FATF digital identity guidance and the upcoming EU AI Act.
- For researchers and academia: Dedicated studies on deepfake detection performance across varied demographic, linguistic, and regional populations of India should be prioritized. Current models are mostly trained on Western data.
- For citizens: Face recordings and other biometric information should be treated with the same caution as sensitive financial details. Be wary of unsolicited video calls, remote interviews, or verification requests from unknown people, and report suspicious activities on any account immediately via the National Cybercrime Helpline (1930) or cybercrime.gov.in.
Conclusion
The I4C advisory of June 2026 marks a critical recognition that advances in generative AI have fundamentally challenged the reliability of facial biometric authentication. For a country whose digital financial ecosystem relies heavily on remote identity verification, the implications are significant. The integrity of India's financial inclusion framework now depends on rapidly strengthening identity assurance mechanisms. Addressing this threat will require coordinated action by regulators, financial institutions, technology developers, researchers, and citizens to develop robust technical standards, enhance detection capabilities, and build public awareness at a pace matching the evolution of AI-enabled fraud.
References and Sources
- I4C / NCTAU Advisory, June 2026 — National Cybercrime Threat Analytics Unit, Indian Cyber Crime Coordination Centre, Ministry of Home Affairs, Government of India. Advisory on AI-Enabled Deepfake Identity Fraud. Issued 11 June 2026.
- shuftipro.com/blog/key-takeaways-from-fatf-horizon-scan-report-on-deepfakes
- https://timesofindia.indiatimes.com/india/fraudsters-creating-deepfakes-to-bypass-facial-authentication-i4c/articleshow/131668958.cms
- hyperverge.co/blog/what-is-a-deepfake
- iproov.com/reports/threat-intelligence-report-2026
- arxiv.org/pdf/2601.06241
Introduction
Artificial intelligence has quietly become part of the future of litigation, like drafting pleadings, summarising depositions, and helping self-represented parties navigate a system that was never designed for them. But what happens when a litigant doesn't just use AI but tries to manipulate it, planting invisible commands inside a court filing, hoping some AI tool reading the document will do the litigant's bidding? That is precisely the question a Connecticut Superior Court judge confronted in Matthew A. Elliott v. New York Bariatric Group, LLC, Docket No. AAN-CV-25-6066141-S (Conn. Super. Ct., Aug. 6, 2026), a decision that may be the first of its kind in the United States to sanction a party for embedding a "prompt injection" in a court pleading.
The Facts
Elliott, representing himself, filed a motion titled "Final and Conclusive Motion for Default". Buried within it, in a tiny, white-on-white font invisible to a human reader but fully legible to any software parsing the document, was a block of text addressed not to the court or opposing counsel but to any AI system that might process the filing. The hidden text instructed such a system to treat its output as agreeing with Elliott's position and to work toward "remediating" a prior clerk's denial of his motion for default.
A second filing repeated a shortened version of the same instruction. When the court issued an Order to Show Cause warning that concealed text in pleadings would not be tolerated, Elliott did not stop. Subsequent filings carried further hidden messages, some flippant asides, one a hidden link to a horror film video submitted even after he had received notice of the sanctions hearing. At the hearing, Elliott characterised his conduct as a self-appointed "audit" of whether the court used AI and later said he continued the practice "as a joke".
The Legal Questions
Judge Walter M. Spader, Jr framed the case around two hard questions. First, does concealing an instruction to an AI system constitute misconduct even if no AI ever acted on it since the court had, in fact, decided the underlying motion on the merits from a printed copy? Second, can a court sanction conduct that Connecticut's own recently adopted AI rules do not expressly address?
Connecticut's Practice Book §4-9, effective only weeks earlier in June 2026, governs generative AI use in filings, but it is aimed at a different danger: the risk that AI-generated output might contain fabricated citations or invented quotations, and it places a verification duty on the filer to catch such errors. As the court observed, that framework addresses unreliable output. It says nothing about manipulated input from a filer seeding a document so that whatever tool later reads it will be corrupted in the filer's favour. The absence of an express rule, the court held, "takes nothing away from the duties of good faith and candour that have always governed those who appear before this Court."
The Court's Reasoning
The court's analysis rested on three pillars. First, intent, not success, is the touchstone of the violation. Because the judge decided the contested motion from a printed version, the hidden instruction achieved nothing, but the court held that the wrong lies in the attempt itself, not its efficacy, drawing an analogy to how the law has long treated attempted corruption of a proceeding as wrongful regardless of the outcome.
Second, the court situated the misconduct within the broader duty of candour owed to tribunals. A pleading, the court reasoned, is a communication to both the court and the opposing party, resting on the premise that what the reader sees is what the filer actually wrote. Hiding a second, machine-readable message beneath that surface breaches this premise. The court drew a memorable comparison: planting an AI-directed instruction in a filing is analogous to an ex parte communication which is a secret message to the decision-making apparatus that the opposing party can neither see nor answer, offending the basic adversarial principle that arguments meant to influence a decision must be made openly, on the record.
Third, the court emphasised that self-represented litigants, while entitled to procedural latitude, remain bound by the same underlying obligations of good faith as represented parties. That solicitude "stops at the misuse of the process itself".
Notably, the court situated Elliott's conduct within a growing pattern well beyond the courtroom, citing reports of job applicants hiding white-text instructions in résumés to manipulate AI screening tools and a professor who caught AI-assisted cheating by embedding a hidden trap word in an exam. Prompt injection, the court noted, has become a documented, catalogued vulnerability recognised across the cybersecurity field, and its migration into litigation was, in the court's words, "unsurprising" given how commonplace the tactic has become elsewhere.
Comparison to Mata v. Avianca
The decision draws a deliberate contrast with the now-famous Mata v. Avianca, Inc. (S.D.N.Y. 2023), where attorneys were sanctioned for submitting briefs citing wholly fictitious cases generated by ChatGPT. Both cases involve AI misuse sanctioned under a court's inherent authority, but the underlying wrongs are different in kind. Mata's concerned negligent reliance on defective AI output; the lawyers there did not intend to deceive the court, and their candour and contrition were treated as mitigating factors even as sanctions were imposed. Elliott's conduct, by contrast, was deliberate input manipulation aimed at corrupting how any AI reader would process his own filing, and it persisted even after a direct judicial warning. As the court put it, "What may have earned a 'no harm, no foul' sanction when it was first done calls for a firmer response when it is done repeatedly after warning."
The court also cited a Brazilian labour court decision, Elisandro Martins de Barros v. Renato Ribeiro de Lima (2026), where two licensed attorneys used a similar hidden-text technique in a jurisdiction where the tribunal actually deployed AI to process filings and where the tribunal's system caught and blocked the injection, followed by a referral to attorney-discipline authorities.
The Sanction and Its Significance
Rather than dismissing the case or imposing monetary penalties, the court chose a narrowly tailored remedy: rescinding Elliott's e-filing privileges and requiring all future filings to be made in person on paper, a sanction addressing the specific abuse (concealed digital text) without barring courthouse access altogether. Importantly, the court reaffirmed that generative AI remains welcome as a litigation aid, provided any output is independently verified, consistent with Practice Book §4-9(b).
Conclusion
Elliott is a small case with an outsized signal: courts are beginning to recognise that AI-era misconduct is not limited to fabricated citations but extends to covert attempts to manipulate the tools, including tools opposing counsel, clerks, or even the court itself might someday rely on. For practitioners, the lesson is to treat every incoming AI-processed document, from opposing productions to client materials, with the same scrutiny once reserved for verifying citations. For courts, it is a reminder that inherent authority over the integrity of proceedings can reach conduct that emerging procedural rules have not yet caught up to naming.
References
Executive Summary
A graphic featuring Union Home Minister Amit Shah is being widely shared on social media, claiming that he has announced a complete ban on alcohol in West Bengal from September 30. The post further suggests that the state will move towards becoming a dry state. Notably, this claim surfaced soon after the BJP’s victory in the West Bengal Assembly elections. CyberPeace Research Wing research has found the viral claim to be false. Our research confirms that Home Minister Amit Shah has not made any such announcement.
Claim:
On Instagram, a user shared a viral graphic on May 8, 2026, alleging that Amit Shah announced a complete ban on alcohol in West Bengal starting September 30. The post link and archived version are provided below:
- https://www.instagram.com/reel/DYDy13zINV5/
- https://archive.ph/mYpZS
Fact Check
To verify the claim, we conducted a keyword-based search on Google. However, we did not find any credible media reports supporting the viral claim. Since the graphic carried the logo of India Today, we also checked the official website, YouTube channel, and social media handles of India Today. However, no matching report or graphic was found.
In the final step, we reviewed the official X account of the Ministry of Home Affairs. Even there, no statement or report confirming the viral claim was found. The relevant link is provided below:
- https://x.com/HMOIndia
Conclusion:
Our research confirms that Home Minister Amit Shah has made no such announcement regarding a complete alcohol ban in West Bengal.
