The Ministry of Electronics and Information Technology (MeitY) has asked Meta to develop its systems to detect and remove deepfakes, manipulated media, propaganda and other illegal or harmful content circulating on its platforms.
The directive follows discussions between government officials and Meta in which Meta admitted gaps in its existing systems and detailed technical solutions being considered to address the growing challenge posed by synthetic and manipulated content.
Officials have asked Meta to develop a clear compliance roadmap for them in which it will outline steps it will take to improve its systems and make it more adhering to India’s regulatory requirements.
Deepfakes and AI-generated content have become a big concern for governments and technology platforms worldwide. In recent years, as generative artificial intelligence has become more and more popular, realistic videos, images and sound can be created in ways that falsely depict people talking or doing things that didn’t happen.
Such propaganda can have severe consequences if manipulated content is used for false information, impersonation, fraud, propaganda, or other unlawful purposes.
According to the report, Meta's technical team told MeitY officials about how the company has already detected and dealt with harmful content in the first place at the end of the week. The company’s existing steps to moderate and respond to illegal material were discussed, the report said.
Government officials then pushed for more specific algorithmic interventions to reduce the reach and circulation of deepfakes, manipulated images and misleading posts.
The government has also called for stronger content moderation and faster action on takedown requests received from authorised agencies. Speed is considered to be of utmost concern as harmful synthetic content can quickly spread on social media and be posted before platforms are able to detect and remove it.
The officials also touched on data-localisation compliance, which has become an important part of India’s digital policy framework.
The government wants platforms operating in India to make sure their systems and processes comply with applicable Indian laws and regulations. More closely linking technology companies and government agencies is also expected to be involved in identifying and respond to harmful content more quickly.
AI-generated content on social media platforms is at the forefront of discussions in the wake of mounting concerns about its use. Deepfakes are particularly difficult to distinguish from actual content for regular users because they involve high-quality video or audio.
A manipulated picture or video can reach millions of users in a short time with the help of algorithms that recommend it. Once that is known, it is even harder to correct the false information.
That’s why authorities have been focusing not only on the removal of posts in question but also on the algorithms and recommendation systems that control how content is shared on the web.
MeitY's demand for algorithmic changes indicates that the government wants platforms to be proactive instead of just reacting to user complaints or takedowns post-publication.
For Meta, these discussions could require more investment in automated detection technologies, human moderation and systems to identify synthetic media before it reaches large audiences.
The company operates a number of major platforms in India such as Facebook and Instagram, and its content policies and recommendation systems have considerable impact on the country’s digital information ecosystem.
Both sides have agreed to continue discussions, and another meeting is expected next week. Meta’s technical plans, compliance roadmap and processes for quicker coordination with authorised agencies could be discussed.
The developments show that India’s regulatory emphasis is increasingly moving towards the technological systems behind online content distribution rather than individual pieces of harmful content.
To do so will be to strike a balance between dealing with illegal and deliberately deceptive material while also promoting legitimate expression and not putting a hold on what some people call legitimate expression.
As content generated using AI is considered more sophisticated, governments and technology companies will have much more pressure to come up with better authentication, detection and accountability mechanisms to prevent and detect threats.
For India, the latest talks with Meta are another step towards the nation’s efforts to strengthen its response to deepfakes and synthetic media. How well the proposed solutions will work as they are implemented will depend on how quickly they are implemented and if they can keep pace with the fast-changing AI technologies.
