Your Feed is Their Fuel
Every day, millions of Indians share moments from their lives online—weddings, birthdays, holidays, and simple everyday joys. These images, posted on public profiles on platforms like Instagram and Facebook, have become a vast, unintended library for artificial
intelligence companies. This process is often called 'data scraping,' where automated bots systematically download publicly available information, including photos and their captions, to build massive datasets. These datasets are the foundation for training generative AI models, teaching them to recognise faces, objects, and scenes, and even to create entirely new, synthetic images. While tech companies like Meta state they primarily use public content, not private messages, this still includes a vast trove of photos and posts from users over 18, sometimes dating back over a decade.
The Legal Shield: India's DPDP Act
The key question for Indian users is: is this legal? India's Digital Personal Data Protection Act (DPDPA), 2023, provides a new framework for data privacy. The law is built on the principle of consent, meaning organisations are generally required to get your explicit permission before processing your personal data. Using personal data for AI training without separate, explicit consent can be a violation of the DPDPA. However, the law also includes exemptions, such as for data that individuals voluntarily make public. This creates a legal grey area. While the DPDPA aims to protect citizens, its application to the vast, automated world of AI data scraping is still being tested and clarified. The government is also tightening rules, requiring platforms to label AI-generated content and take down harmful material like deepfakes within hours, signalling a move toward greater accountability.
Beyond Privacy: Understanding the Risks
The consequences of your photos being used in AI training go beyond a simple loss of privacy. One of the most significant dangers is the creation of 'deepfakes,' which are realistic but fake images or videos of a person. Your likeness could be used without your permission in scams, misinformation campaigns, or even harassing or explicit content. Furthermore, the data from your photos, when combined with other information, can be used to build detailed profiles for targeted advertising or, in more sinister scenarios, for surveillance. Even if an AI app seems harmless, like a fun filter trend, it might be collecting your biometric data. Once your face is part of a training model, it's incredibly difficult to remove, as the AI has already 'learned' from it.
Practical Steps for Digital Protection
While the landscape is complex, you are not powerless. The most effective step is to manage your privacy settings. Making your social media accounts private is the strongest defence, as it prevents your content from being scraped by third parties. For public accounts, especially for creators, be selective about what you post. Avoid sharing high-resolution images of family members, particularly children. Many platforms are now offering specific opt-out settings for AI training; take the time to find and enable them. Consider using visible watermarks, though be aware that sophisticated AI can sometimes remove them. For added protection, technical tools like Glaze are emerging, which can 'cloak' images by making subtle pixel-level changes that confuse AI models trying to learn an artistic style. Finally, talk to your friends and family about their posting habits, especially concerning photos of you.
