Centre asks Meta to act against deepfakes, probes role of algorithms

The Centre has asked Meta to take corrective measures against deepfakes and other harmful AI-generated content on its platforms, news agency PTI reported on Saturday. The government is also examining whether Meta continues to qualify as an “intermediary” under Indian law, particularly given the role of its recommendation systems in deciding what content users see.

According to PTI, the government is also considering seeking legal opinion on further action against Meta after assessing the outcome of its recent meetings with the company.

The review could subsequently be expanded to other digital platforms to assess whether their recommendation systems and content practices comply with the legal requirements for intermediary status.

The government has said its objective is not censorship but ensuring compliance with Indian laws. Meta, whose global affairs head Joel Kaplan met IT Minister Ashwini Vaishnaw and IT Secretary S Krishnan earlier this week, has acknowledged “serious issues” and assured the government that it is working to address the concerns.

Why Meta’s intermediary status is under the lens

At the heart of the government’s scrutiny is the role of “recommendation algorithms,” which determine the content shown to individual users and could have a bearing on the legal protections available to platforms.

The PTI report cited government sources as saying that the meetings with Meta have raised questions over whether its recommendation systems and paid content promotion are consistent with its status as an intermediary under the Information Technology Act, 2000. The key issue is whether a platform that actively decides “what content is shown to whom” could be seen as moving beyond the role of an intermediary and into that of a publisher.

The issue is significant because Section 79 of the IT Act provides intermediaries with 'safe harbour' protection from liability for third-party content, subject to compliance with prescribed conditions and due diligence requirements. These obligations are set out in the IT Act and the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021.

If Meta’s systems determine what content users see and also promote content for payment, questions could arise over whether these functions are compatible with the legal definition of an intermediary. Failure to comply with due diligence requirements can result in the loss of the exemption from liability under Section 79, besides other action under applicable laws.

Centre flags deepfakes, AI-generated content

The government has also raised concerns over the continued circulation and virality of harmful AI-generated content, including deepfakes, even after such material has been flagged. Officials questioned Meta over why AI-generated videos without appropriate labels continue to appear on its platforms despite rules requiring synthetic content to be identified and labelled.

During two days of discussions this week, Meta’s global team was questioned on deepfakes, child sexual abuse material (CSAM), unlabelled synthetic content, recommendation systems and the safeguards used to ensure compliance with Indian laws. The company held technical discussions with officials on measures to address these concerns.

The government has asked Meta to strengthen human oversight in content moderation and improve its understanding of Indian languages and local contexts. Officials are expected to follow up regularly, with another meeting likely next week.

The meetings followed the temporary restriction of Prime Minister Narendra Modi’s Facebook post, for which Meta apologised. A Parliamentary Standing Committee had also sought an apology from Meta chief executive Mark Zuckerberg, warning that the legal protections available to the company could be reconsidered.