Meta more cautious on CSAM after unequivocal govt message; action visible: IT Ministry sources

The government's hard talk with Meta has led to action on child sexual abuse material (CSAM), with the company now showing greater sensitivity and caution on such content, IT ministry sources said on Tuesday, asserting that the Centre has made it unequivocally clear that such content constitutes a violation and cannot be covered by safe-harbour protection.

Sources said the government's intent is not censorship or restricting coverage in any way, but ensuring Indian laws and norms are followed and that social media harms are mitigated.

Meta explained how its algorithms and systems work, the ministry sources said, noting that discussions are on and such matters involve "judgment calls" and "tweaks".


The IT Ministry official further said that Meta has shown increased sensitivity and caution on CSAM, with the government seeing some action by the company following its recent discussions. Sources described the response as "important and significant".

The government had unequivocally conveyed to Meta that CSAM violates Indian law and that no safe-harbour protection is available for content that violates the law in such a manner, sources said.

They added that the message appears to have "gone home", with such content not resurfacing so far and the company being more cautious on this front.

Sources clarified that safe harbour "removal" per se is not done by an act of government, but in the court of law. Violation of Section 67 of IT Act can lead to prosecution.

Asked about the issue of algorithmic bias, the official said: "We are discussing about their algorithms and how they work and what happens. They have been explaining how these things work and how the structure actually operates and functions."

WhatsApp username row

On the WhatsApp username row, sources said the matter is under examination as it doesn't involve just that particular platform, but other messaging apps as well.

"So we are in the process of examining (the issue)", sources said.

It may be recalled that WhatsApp's proposed username feature, which would allow users to communicate without sharing phone numbers, had come under government scrutiny over concerns that it could increase risks of online fraud, phishing, digital arrest scams and impersonation. The IT ministry had asked Meta to explain why action should not be initiated under the IT Act and rules, and had directed WhatsApp not to launch the feature in India until consultations are completed to the Centre's satisfaction.

WhatsApp, which has about 50 crore users in India, subsequently submitted its response and assured the government that the feature would not be rolled out until discussions conclude. The IT ministry has also issued notices to Telegram and Signal over their existing username features, seeking details on safeguards against fraud and impersonation.

On the talks with Meta, the ministry sources said the US tech giant has indicated it is working to prevent content already taken down (as unlawful) from repeatedly resurfacing, including through improvements to its filters. On the government's call for greater sensitivity to Indian cultural and local context in content moderation, Meta has also indicated that it is taking action, the source said.

"We have to continue to watch, and these are things which overnight you can't say I'm fully satisfied with what they've done. We have to actually look at what really happens and what we see happening on social media," the source said.

The government's objective is to ensure Indian laws are followed and platforms do not create situations which are adverse to the country's social norms, while striking the right balance between the benefits platforms provide and risks, say, of addiction, gaming-related harms and CSAM.

"All of these have to be addressed and clearly they (platforms) have a responsibility," the source said.

These concerns are not unique to India and several democracies, including Australia, the UK and France are grappling with it, they said, adding that platforms have a responsibility to mitigate such harms while recognising differing cultural contexts.

Sources said the government has acted within India's constitutional and statutory framework, including through Section 69 A of the IT Act, which it has used sparingly where prescribed conditions are fulfilled.

AI labelling

On AI labelling, the ministry official said the law is clear that synthetically generated content must be labelled, and that responsibility lies at three levels, including with platforms. Platforms must comply accordingly, they added.

Meta's team led by its global affairs head Joel Kaplan had met IT Minister Ashwini Vaishnaw and IT Secretary S Krishnan earlier this month, following the government's summons over the temporary restriction of Prime Minister Narendra Modi's Facebook post. In the rounds of discussions that followed, the Meta team apologised, and was asked by the government if it is adhering to Indian laws, grilled on algorithmic issues and guardrails meant to ensure that constitutional and legal requirements are being met.

Meta had assured that it is indeed serious and "working hard" to address concerns around deepfakes, child sexual abuse material, bots and synthetic content.