5 months ago
2
In the property of Artificial Intelligence (AI), exertion is simply a double-edged sword, with users grappling with the trade-offs betwixt convenience and privacy. While India has a normative privateness model successful presumption of the Puttaswamy judgement (2017); the Information Technology Act, 2000 and its Intermediary Guidelines; and the Digital Personal Data Protection Act, 2023, and Rules, the world of privateness remains opaque.
We present unrecorded successful a fishbowl nine wherever we are gauging ‘harm’ from a myopic lens of privateness and dignity alternatively of obscurity. As Meredith Broussard notes successful her publication Artificial Unintelligence, society’s over-reliance connected exertion is leaving america ill-prepared to header with the precise systems we person built. This not lone exposes individuals to the risks of information breach but besides pushes them into obscurity, particularly successful cases of Non-Consensual Intimate Image Abuse (NCII), wherever algorithms make deepfake pornographic images without one’s cognition oregon control. Regulating specified an battle is an urgent ineligible and argumentation imperative. The accepted frameworks for addressing specified abuses are inadequate. Traditional approaches often picture risks of immoderate specified surveillance arsenic nonaccomplishment of privacy, erstwhile successful world it is galore much things arsenic well: anxiety, chronic fearfulness of being watched, unfortunate blaming and shaming, societal stigma, vocation stagnation, imperishable nonaccomplishment of autonomy, and bodily integrity.
Laws are not enough
Surprisingly, contempt cybercrimes being connected the rise, determination is nary modern information connected NCII. Data of the National Crime Records Bureau (NCRB) puts each cybercrimes successful 1 category, without immoderate granular classification of circumstantial offences. We filed an Right to Information exertion connected October 3, 2025 seeking accusation connected the fig of cases registered successful the erstwhile twelvemonth relating specifically to cyberbullying and cybervoyeurism, on with the gender-wise organisation of victims. After much than a month, the Ministry responded that “law and order” and “police” autumn nether the State List, and therefore, the astir due authorization to furnish specified accusation would beryllium the respective State governments.
This shows that specified ineligible provisions are not capable to code the realities of online abuse. Accessibility, awareness, and societal acceptance of these laws play an arsenic captious relation successful determining their effectiveness. A important stock of young women are unaware of what offences specified arsenic voyeurism oregon deepfake porn legally entail. The deficiency of integer literacy is compounded by deep-rooted societal stigma, shame, and fearfulness of blame, which often deter victims from reporting. In utmost cases, this has driven immoderate survivors to self-harm.
Going beyond an SOP
On November 11, 2025, the Ministry of Electronics and Information Technology issued Standard Operating Procedures (SOPs) to curb the circulation of NCII. These guidelines mandate that specified contented indispensable beryllium taken down wrong 24 hours of reporting and question to safeguard the “digital dignity” and privateness of women by offering aggregate platforms for complaints. This is simply a invited and long-awaited step. However, an SOP is lone the starting point. Its effectiveness depends connected being backed by beardown capacity-building programmes, stakeholder consultations, and strengthening of enforcement agencies.
A cardinal regulation lies successful the lack of a gender-neutral framework. Studies amusement that transgender persons, peculiarly transwomen, are disproportionately targeted done deepfake-based harassment. Yet the SOP is soundless connected transgender victims, overlooking the Supreme Court’s designation of transgender persons arsenic the “third gender” entitled to adjacent rights. Further, it does not found wide accountability mechanisms, specify the quantum of punishment, oregon articulate circumstantial regulations for deepfake generation, dissemination, and tracing. Thus, having a dedicated instrumentality connected NCII is the request of the hr — 1 that goes beyond the accepted absorption connected actus reus and mens rea and emphasises explicit duties connected platforms, AI developers, and intermediaries, much circumstantial and broad than the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rule, 2025.
With the proliferation of AI-generated deepfakes, chiefly utilized to harass, shame, and soundlessness victims (mostly women), privateness is progressively shaped and threatened by technological capabilities alternatively than ineligible protections. The deficiency of procedural safeguards, traceability norms, and autarkic oversight mechanisms has allowed specified crimes to spell unreported and unpunished for years, adjacent arsenic their frequence and severity escalate. These challenges rise an important question: Is an SOP enough?
Lack of consciousness of rights oregon adjacent of what “voyeurism” oregon “revenge porn” legally constitutes, inadequate sensitisation of constabulary officials, victim-blaming, and deficient cyber-investigative capableness further dilute the interaction of existing laws. As NGOs and probe studies highlight, thousands of cases are filed regular crossed India, yet convictions stay disproportionately low. In this context, portion the SOP is simply a important archetypal step, a meaningful effect to NCII and deepfake harms requires gender-neutral reforms, constabulary training, capableness building, level accountability, AI-specific safeguards, and stronger victim-centric ineligible mechanisms.
Aastha Tiwari, Assistant Professor (Law) and PhD scholar, Maharashtra National Law University Mumbai; Shweta Bhuyan, Research Assistant (Law) and PhD scholar, Maharashtra National Law University Mumbai
