Draft Rules Detailed Unpacked

The draft inserts a new statutory definition for “synthetically generated information.” Moreover, it introduces deepfake identification mandates that demand non-removable labels or metadata. Visual items must display a 10 percent badge. Audio clips need an audible cue during their first 10 percent. Additionally, large platforms must verify user declarations before publication. These steps extend existing IT ministry proposal duties under the 2021 Intermediary Rules. In contrast, earlier rules lacked specific marking thresholds. The text also clarifies safe-harbour for good-faith removals, strengthening takedown oversight incentives. Finally, explicit transparency requirements call for visible, user-friendly notices alongside hidden provenance tags.

These measures promise consumer clarity. Nevertheless, experts warn that broad definitions may capture satire or benign edits. Therefore, precise guidance will be vital. Next, we assess platform obligations in detail.

Platform Duties Intensely Heighten

Significant Social Media Intermediaries (SSMIs) with over five million users shoulder the heaviest burden. Firstly, they must compel uploaders to self-declare synthetic status. Secondly, automated tools should confirm claims, fulfilling the IT ministry proposal verification clause. Furthermore, failure triggers potential safe-harbour loss. Meanwhile, generator services that “offer a computer resource” must embed identifiers at creation. Such obligations mirror China’s regime yet adopt India-specific metrics.

Platform counsel highlight three friction points. • Ex ante detection accuracy remains unreliable. • Interface changes may hurt user experience. • Global distribution complicates consistent watermark display. Consequently, compliance spending will climb. These concerns also intersect with takedown oversight expectations, because missing labels may force removals. Moreover, fulfilling deepfake identification mandates while preserving creative quality appears challenging. Nevertheless, clear transparency requirements could boost user trust.

Heightened duties expand accountability. However, technical feasibility questions persist. The next section reviews stakeholder sentiment during the public feedback period.

Industry Reactions Remain Mixed

Government officials frame the move as transparency, not prohibition. Secretary S. Krishnan stressed, “Label, don’t restrict.” Conversely, platform representatives request flexible standards. Additionally, legal firms argue the “reasonably appears authentic” test is subjective. Civil-society groups applaud stronger takedown oversight but fear over-broad enforcement. Meanwhile, creators worry the 10 percent watermark may jar audiences. Independent think-tanks urge phased rollout of deepfake identification mandates.

During the ongoing public feedback period, stakeholders must submit view points by 13 November 2025. Submissions may influence scope, timeline, and transparency requirements. Furthermore, some commentators ask MeitY to publish reference watermark schemas. Consequently, final text could tighten or relax obligations.

Reactions reveal diverging priorities. Nevertheless, comparative insights offer broader perspective, as discussed next.

Wider Global Policy Context

India joins a small group of jurisdictions mandating provenance cues. The EU AI Act outlines similar transparency requirements, although enforcement paths differ. China already enforces visible watermarks and metadata tags. Meanwhile, the United States relies on voluntary policies. Therefore, the IT ministry proposal positions India as an early, prescriptive adopter.

In contrast to EU guidance, New Delhi specified numeric visibility—10 percent—inside its deepfake identification mandates. Moreover, India’s safe-harbour proviso attempts balance by shielding compliant takedowns. Consequently, policy analysts predict ripple effects across regional markets. Platforms may harmonize disclosure flows to serve multiple regulators, reducing marginal costs.

International precedents highlight alignment opportunities. However, local implementation hurdles deserve close attention.

Major Compliance Hurdles Ahead

Verification accuracy tops engineering concerns. Automated detectors misclassify content, risking wrongful flags and extra takedown oversight. Additionally, non-removable metadata may expose creator identities, raising privacy issues. Moreover, adversarial actors can strip or obscure labels, undermining deepfake identification mandates. Cross-border hosting further complicates enforcement.

Smaller startups fear resource strain. Consequently, they might restrict generative features within India. Meanwhile, creative professionals argue large watermarks hinder aesthetics, conflicting with the proposal’s transparency requirements. Nevertheless, compliance also offers benefits—clear guardrails could unlock advertising confidence and reduce legal uncertainty.

These hurdles sharpen implementation debates. However, strategic planning can ease the transition, as the next section outlines.

Practical Strategic Next Steps

Enterprises should map content workflows immediately. Firstly, audit generation tools for embedded identifier capabilities. Secondly, prototype visible labels meeting the 10 percent rule. Thirdly, refine policy documentation before the public feedback period closes. Additionally, invest in robust detection pipelines to satisfy the IT ministry proposal. Meanwhile, staff must train on prompt takedown oversight procedures.

Professionals can also strengthen competencies through accredited learning. For instance, ethical leads may pursue the AI Ethics Strategist™ certification. This programme covers governance frameworks, risk mapping, and transparency requirements alignment. Consequently, organisations gain internal expertise to navigate deepfake identification mandates.

Strategic steps create readiness. Nevertheless, final rule wording will confirm exact obligations, so continuous monitoring remains essential.

The proposal’s journey now enters a decisive phase. Therefore, proactive engagement will shape workable, responsible AI governance.

Key Figures Snapshot

Meanwhile, consider the scale behind the draft:

• 1.03 billion Indian internet users (Oct 2025)
• 500 million social-media identities across platforms
• SSMI threshold: 5 million registered users
• Consultation deadline: 13 November 2025
• Label size rule: 10 percent visual/audio duration

These figures illustrate the vast reach of synthetic media risks. Consequently, they justify MeitY’s assertive stance.

Statistics anchor the discussion. Next, we conclude with overarching insights and actions.

Conclusion And Call-To-Action

India’s IT ministry proposal seeks to demystify synthetic content through labels, metadata, and stronger platform duties. Moreover, it intertwines deepfake identification mandates, rigorous takedown oversight, and explicit transparency requirements. Nevertheless, practical challenges around detection, cost, and creative freedom persist. Furthermore, the ongoing public feedback period offers the last chance to refine details.

Therefore, organisations should analyse workflows, file considered submissions, and upskill governance teams. Professionals ready to lead can explore the AI Ethics Strategist™ credential. Engage today, influence tomorrow, and help craft responsible AI futures.