By Nantoo Banerjee
Fake news is flooding across India’s nearly half-a-billion social media and messaging platforms, threatening trust, relationships, and democratic processes. The country, boasting almost a billion internet subscriptions, is facing a growing crisis of misinformation with digital content spreading faster than truth across the online community. To add to the concern are Artificial Intelligence (AI) generated images accelerating the spread of digital misinformation by creating highly realistic but completely fabricated visuals that strengthen fake headlines. With AI tools producing these synthetic images cheaply and at scale, fact-checkers struggle to keep up, particularly during emotionally charged breaking news cycles and global conflicts, making audiences more likely to believe and share unverified claims. Seemingly realistic and “probative” images—those that provide strong visual evidence to false headlines—significantly boost a reader’s belief in the misinformation.
Fabricated visuals of politicians, and unverified crisis footage frequently go viral across largely unregulated social media platforms, often before fact-checking algorithms can intervene. The AI technology has fuelled beyond political and current events a massive surge in non-consensual deepfakes and manipulated images targeted at individuals, particularly women. Media scepticism is undermined by media cynicism. While encountering suspicious media, few care to pause and utilize tools like reverse image searches to verify the source’s origin before sharing. Not many net users spend time to understand how these images are created and identify common digital artifacts to act as effective interventions. Interestingly, a 2024 Pew survey showed as many as 65 percent of Indians consider fake news a very big problem, while 81 percent saw it as at least moderately concerning.
AI-generated deepfakes now pose an urgent global security and social trust crisis. Fuelled by cheap, accessible AI tools, synthetic media—including video face-swaps and voice clones—are heavily weaponized to commit large-scale financial fraud, manipulate elections, and generate non-consensual sexualized imagery targeting women and children. The proliferation of these digital forgeries undermines public discourse and makes it increasingly difficult for the public to discern fact from fiction. Because they are so convincing, deepfakes threaten organizational reputations and contribute to a broader atmosphere of doubt, where genuine evidence can be falsely dismissed. Several organizations and governments are actively fighting back by ramping up protective measures. Regulatory watchdogs advise deploying real-time detection programs, applying digital watermarks, and shifting to a “zero-trust” authentication mindset.
While the pervasive nature of synthetic media has triggered widespread legislative and regulatory crackdowns, India’s effort to enforce strict guidelines for social media intermediaries is highly laudable – especially, the latest statement by MeitY (Ministry of Electronics and Information Technology) secretary Sundaram Krishnan, suggesting that the government is looking into a separate legal framework for AI. Earlier this month, Krishnan told a business gathering in Delhi that India has been relying on a combination of IT rules, the Digital Personal Data Protection (DPDP) Act, and criminal codes like the Bharatiya Nyaya Sanhita (BNS) to manage AI. The proposed new legislation aims to proactively regulate emerging risks and cybersecurity threats while balancing ongoing AI innovation.
The industry as well as the public are concerned. Obviously, the existing broad based legal framework and intermediary rules are not sufficient to address AI-related issues such as deepfakes, misinformation and online harms. The rapid proliferation of generative AI tools has made deepfakes, misinformation, and online harms critical issues across the tech industry, prompting governments and international organizations to step up regulations. These advancements present a variety of tangible threats such as disinformation, financial fraud and extortion and personal harassment. Synthetic video and audio clips are increasingly used to manipulate public discourse and interfere with democratic processes. The United Nations (UNHCR) warned that AI-generated deepfakes are spreading life-threatening misinformation to displaced populations.
In India, cybercriminals regularly use voice cloning and video synthesis to impersonate executives and family members, especially pensioners and elderly, facilitating wire transfer fraud and extortion. Deepfake technology is being frequently weaponized to create non-consensual intimate imagery or targeted defamatory content. Platforms such as X and Instagram are required to take down identified deepfakes within three hours. The government proposes stricter continuous-labelling mandates for synthetically generated content and exploring a dedicated AI-focused legislative framework to manage these systemic risks. Globally, the International Telecommunication Union and Global AI dialogue continue to push for incorporating concrete human rights safeguards across the AI life cycle.
Lately, over 60 global data protection and privacy authorities have issued joint statements warning against the unauthorized spread of synthetic imagery of real individuals, demanding stronger safety-by-design guardrails from AI developers. International telecommunication and tech bodies are actively developing standardized watermarking and provenance tracking to help the public verify the authenticity of digital content before consuming it. Several global powers have established unique AI policy frameworks to balance innovation with safety, security, and human rights. The world’s primary regulatory approaches are led by the US, the European Union, China, and rising technology hubs like India, Singapore, and South Korea. The US uses sector-specific guidelines and executive actions rather than a single, sweeping federal law, to frame its AI policy.
The European Union’s approach is toward a comprehensive, risk-based legislative framework. The landmark EU AI Act classifies AI into different risk categories (from unacceptable to minimal risk). It strictly bans applications like social scoring, biometric categorization of sensitive traits, and untargeted facial recognition in public spaces. China follows state-centric regulations targeting the management and control of AI-generated content. It implemented administrative regulations targeting generative AI services and rules on algorithmic recommendations. These policies require providers to ensure AI outputs reflect “core socialist values,” avoid subverting state power, and explicitly label AI-generated content. South Korea has a comprehensive national safety law combined with aggressive support for AI research and development. Its AI Framework Act is one of the world’s first comprehensive laws regulating AI safety, emphasizing transparency and promotional measures to support small-to-medium enterprises and workforce preparation.
Until now, India has been focusing heavily on integrating AI to drive socioeconomic development, digital infrastructure, and national welfare. The IndiaAI Mission and various advisory frameworks highlight ethical use, bias mitigation, and the responsible deployment of AI in governance systems, supported by significant government and startup funding. The IndiaAI Mission is a comprehensive Rs. 10,372 crore initiative approved by the government to build a sovereign AI ecosystem for the country. Implemented by IndiaAI under MeitY, the mission aims to democratize computing access, build indigenous foundation models, and foster socially impactful AI solutions. It is encouraging to note that IIT Kharagpur, IIT Madras, IIT Jodhpur and IIT Mandi have been working on technology and forensic frameworks to detect deepfakes, digital fraud, voice cloning and counter AI-generated manipulation and improve online security. (IPA Service)
