Govt's AI advisory only for large platforms, not for startups: MeitY

Union minister Rajeev Chandrasekhar said that the advisory was aimed at preventing untested AI platforms from deploying on the Indian internet

Union Minister Rajeev Chandrasekhar on Monday said the recent advisory for artificial-intelligence (AI) platforms, issued by the Ministry of Electronics and Information Technology (MeitY), applied only to large ones and not for startups.

While briefing the media on the advisory, Chandrasekhar said it was sent to all AI platforms and intermediaries in the country, including Google Gemini and Ola Krutrim, India’s first AI unicorn.

Ola Krutrim has said it will adhere to the guidelines.

Chandrasekhar said the advisory aimed to prevent untested AI platforms from being deployed on the Indian internet.

“Process of seeking permission, labelling and consent based disclosure to users about untested platforms is an insurance policy to platforms who can otherwise be sued by consumers,” he further added.

MeitY’s advisory has asked social-media intermediaries and AI platforms to get permission before launching AI products in the country.

The advisory stated companies hosting unreliable or under-testing AI platforms and wishing to create a sandbox on the internet for testing must get permission from the government and should label the platform “under-testing”.

In another post Chandrasekhar said: “Advise to those deploying lab level /undertested AI platforms onto public Internet and that cause harm or enable unlawful content - to be aware that, platforms have clear existing obligations under IT and criminal law. So best way to protect urself is to use labelling and explicit consent and if you are a major platform take permission from govt before you deploy error prone platforms.”

“Indian Internet is governed by laws to keep it Safe & Trusted.”

An email sent to Google, Meta, and Snap remained unanswered.

The ministry has asked platforms to ensure that the biases arising from their AI models or platforms do not hamper the electoral process in India.

“Public Internet should not be conflated with a Sandbox in so far as unlawful content is concerned,” said the minister in his post.

All platforms concerned were asked to submit an action-taken cum status report to MeitY within 15 days of the advisory.

According to the Confederation of Indian Industry, this is a welcome step.

“However, an appropriate balance is needed between regulation and allowing innovation for harnessing the true potential of AI,” said a spokesperson.

Industry executives said this might have more impact on social-media platforms than on e-commerce firms.

“E-commerce companies use data related to commerce, products and customers, while social-media platforms including Google and Facebook use diversified data and have to be more cautious now,” said an industry executive.

“However, e-commerce companies may also have to show their large learning models to the government.”

Industry executives said the trigger for the advisory might be Google’s Gemini chatbot, which is facing criticism in India due to a response it generated for Prime Minister Narendra Modi. 

“The other reason is that the government wants to keep control on a lot of deep fakes that may have an impact on the elections. They want the social media platforms to take some responsibility,” said an executive.

The queries to Amazon and Flipkart remained unanswered. 

Perplexity AI Chief Executive Officer (CEO) Aravind Srinivas on X said: “Bad move by India.”

Abacus AI’s CEO Bindu Reddy posted on X: “If you know the Indian government, you know this will a huge drag!”

Vaishnaw defends AI advisory

Union Minister Ashwini Vaishnaw on Monday said the advisory issued by the government is not legally binding and is a step to ensure that Artificial Intelligence (AI) models are properly trained before their public deployment. 

“The advisory is not a regulatory framework, but an advisory to ensure that intermediaries and AI platforms test their models properly before deploying them, so as to ensure the safety and trust of Indian citizens,” said Vaishnaw while speaking at a press conference.

Vaishnaw said that the guidelines will apply to social media platforms, and not to all sectors. 

He mentioned entities working on AI in agriculture and health­ care will not be affected by the advisory. The minister also said that platforms can’t excuse themselves from their responsibility for the biases arising from unreliable and undertested platforms. 

“People come and say sorry we didn’t test the model enough. That is not right, these platforms will have to take the responsibility for what they are doing,” he added. Ashutosh Mishra