Surveillance Reform and India’s new Telecom Act

The recent Telecom Act in India has sparked concerns over potential surveillance implications, especially for their lack of surveillance reform. I was on CNBC-TV18 with Ashmit Kumar towards the end of the year, discussing the act (link in the comments). I’ve had some time to think about this more, so a few points: :

1. Unchecked Power to do surveillance: The Telecom Act lacks sufficient checks on government surveillance, enabling unrestricted access to personal communications, which is a privacy concern. Government officials both issue and evaluate surveillance orders, creating a situation where the entity in power polices itself, leading to a clear conflict of interest.

2. Ubiquity of requests for call data records: Call Data Records (CDRs) are obtained by local police in almost 90% of cases by local police without adequate oversight, compromising the privacy of countless individuals.

3. Intrusive CMS: The Centralised Monitoring System, which is embedded in telecom networks, enables extensive monitoring capabilities of non-WhatsApp calls and messages, without necessary limitations, potentially leading to abuse under the guise of security.

4. Lack of proportionality in surveillance: As per previous reports, around 300 orders per day for surveillance were issued. How can there be adequate application of mind? Also, there needs to be a graded approach: under what circumstances, and based on what kind of information, and what kind of threat, what level of surveillance and for how long, is permitted. As of now, there’s no such limitation.

5. Lack of definition of National Security: There is no definition in law, in terms of what National Security means. It could mean whatever the government wants it to mean. Can snooping on an opposition politician during elections be seen as a national security issue? Currently, it can.

6. Ambiguities in Legislation: The Telecom Act’s vague language could extend its reach beyond telecom networks to the internet, impacting digital privacy. If applied to the internet, the Act could validate IT Rules, which have been challenged by WhatsApp for undermining end-to-end encryption, vital for private digital communication. The minister has clarified that online is not impacted, but the wording of the law opens it up to this kind of interpretation.

How do we fix this?

1. Take the internet out of the ambit of telecom surveillance
2. Bring in Parliamentary and Judicial oversight of surveillance agencies.
3. Create a structure of proportionality for surveillance
4. Enhancing Transparency: Mandating detailed reporting on surveillance orders and their rationale, and only a surveillance that is done under high threat perception of national security may be kept out of oversight of courts.

I don’t expect surveillance reform from the executive, but there is an opportunity that courts have, in terms of addressing the Pegasus case which has been pending in the Supreme Court for a while now.

Watch the show here:

On 28% GST in “real-money gaming” in India

India has levied 28% GST on the entire amount in case of payments made for gaming and real-money-gaming. It’s a controversial decision, with lobbyists for the Real Money Gaming industry, and some astro-turfing consumer group criticising this development. A comprehensive overview here on MediaNama.

I’ve spoken publicly about this issue:

  1. The Core podcast with Govindraj Ethiraj: Listen
  2. CNN News18: Watch
  3. Russia Today

My take:

  1. The online realmoney gaming industry has essentially been treated equally with gambling, when it comes to taxation. In my opinion, they’re not very different, and many countries across the world regulate online real money gaming in the same way that they regulate gambling.
  2. Lobbying over the last two years has focused on driving a distinction between real-money-gaming and gambling, and even sought to treat is as a type of gaming. Even now, comments from lobbyists suggest that “gaming will be destroyed” by this move. In the past, they’ve tried to position themselves as “eSports”, which failed. What irks me is this false equivalence with gaming: Gaming isn’t dependent on people playing to earn: it’s dependent on people playing to win. There’s a difference.
  3. India’s Ministry of Electronics and IT recognised them as online real-money gaming intermediaries even though:
    • The IT Act doesn’t allow creating these new definitions.
    • Gaming isn’t an intermediary function: there are controls and gameplay, as well as algorithmic behavior, that are determined by the Platform. Traditionally, games are traditionally “publishers”, and creators of entertainment. To the extent, they’re a new form of entertainment, because the environment responds to your behavior.
  4. MEITY seems to have jumped through hoops to twist definitions and processes to accomodate the online realmoney gaming industry as a gaming industry under the IT Act, in the IT Rules. There’s a major case in TN about jurisdiction, and gambling is regulated by states. Given the Levels of addiction, TN passed a bill to ban online realmoney gaming/gambling. Telangana hasn’t been too keen on them either. Hence the forum shopping with MEITY, where they’ve found favour, including with MoS for IT, Rajeev Chandrasekhar. I mean, if India wants to legalise gambling, let’s do that. I don’t think MEITY’s “creativity” in dealing with this issue reflects well on those involved.
  5. Over the past year, we’ve seen advertisements without disclaimers about addiction, and promises of making multiples of money, which is an approach similar to gambling. The industry is worried that now, given that consumers are likely to win less, they’ll stop playing…there’s something deeply flawed about an industry that says that their users won’t play if they won’t make enough money when they win.
  6. Credit is due to India’s GST Council and Finance Ministry that they’ve seen through these shenanigans. Clearly there is a worry that this segment is too hot, and its clearly a sin segment. Addiction, indebtedness with payday loan apps, suicides, and they needed to open a valve and release some pressure. Remember crypto? Same kind of gambling situation, with promises of great returns. What did finmin do? In hindsight (and I was very critical of it initially), TDS was a pressure release. Around 60% drop in transactions in the first month. 28% GST will have the same effect on the so called “online realmoney gaming” industry. This is a welcome move. It’s good for consumers. Of course, there are court cases on, both around jurisdiction (state vs centre, gaming vs gambling ). There might be new cases around taxation. Watch this space.

TV Show: The impact of low cost handsets on India’s Internet

On CNBC-TV18, we discussed yesterday the launch of a low cost Reliance Jio handset and its impact on India. My key topic points:

My notes for the show:

  • Handsets are important: Low cost handsets are important for Internet access because they bring access to the internet for users. You cant give 1GB a day for a few hundred rupees a month, and then expect a user to pay Rs 10,000 for a handset. Users also transition eventually from low cost devices to higher cost devices as their understanding of the utility of the Internet increases.
  • India’s internet user base has mostly been stagnating over the last few quarters. So another price war might actually help shake things up a bit.
  • Jio 5G: 6,024 Cities/Towns, but 51 in J&K, 74 in Uttarakhand, 11 towns in Mizoram, 19 in Nagaland, 18 in Manipur
  • Address shared access in low income households: Lower cost handsets also drive personal Internet access, rather than shared Internet access. RIght now many handsets are essentially household internet access, and this makes it difficult for women in these households to get to use the Internet. Lower cost devices address this gap.
  • Downward pricing pressure for other operators: the pricing of the bundle brings Internet pricing down to Rs 123 (around $1.5) per month. Other incumbent mobile operators have increased Internet access pricing recently, and this will probably bring costs down for all users, and other operators feel downward pricing pressure.
  • Great for Startups targeting Bharat: Augurs well for startups targeting Digital Bharat, including agritech startups
  • Jio has been critical for Internet growth in India: if you check the charts, the launch of Jio in 2016 led to a decline in 2G connections and a massive increase in mobile broadband (3G+4G connections). Most people leapfrogged 2G and 3G and went straight to 4G.
  • Ubiquitous infra rollout still needed: We still don’t have adequate 4G infra in all spaces. There are areas in the hills and in between towns and cities where it’s not adequate. At times, in some areas, one mobile operator has better infrastructure than the other, including in case of Jio. I think eventually the availability of network will be more important than pricing of Internet of access in some regions, and that’s where Jio still needs work.
  • Bundling services into the same price and replacing other forms of entertainment also makes a lot of sense. It’s not as if it has not been tried before, like Airtel had Wynk and its own movie services as well… all of this competition is fantastic for consumers.

TV Show: Regulating financial influencers

The influencer ecosystem has been gaining importance in India: they’re better than most media and experts at understanding consumer needs and packaging their content. Many of them are experts at what they do. They monetize their work through advertisements, product placements, and subscription based premium content or advisory models.

They thus replicate many of the functions of media publications.

I was on News9 to discuss a few controversies around the financial influencer space, now referred to as “finfluencers”, and frankly, as long as they don’t play in a regulated space such as stock advisory services, without the required regulatory certification/registration/license met, I think they’re providing an important service of financial education.

I also spoke about the upcoming Digital India Bill in the first part of the show.

More in the video:

Some points from my pre-show preparation about the Digital India Bill:

  • I’d be surprised if the digital India bill is tabled in 15 days. Mr Chandrasekhar a month and half ago said it will be put up for public consultation on June 7th. The dates have been pushed.
  • Very dangerous for Startup India, if safe harbor is removed. Majority of Indian and global companies will not be able to survive
  • The IT Act amendments in 2008, and subsequent court proceedings gave us several freedoms and we should not lose them.
  • I hope we don’t see the revival of 66A
  • Sectoral regulation will hurt innovation: companies take time to find product market fit.

TV Shows: Addressing job loss fears around AI

Are AI tools a boon or a bane. On TV shows and on conference panels, I’ve spoken about how AI tools actually make our work easier, and more efficient. Whether an organisation wants to do more with the greater efficiency or make people redundant is their decision, and no fault of the AI tools. It’s important for people to learn these tools, and add these skill sets.

Two TV shows:

My show notes from the Jan 2023 discussion:

  1. Here’s where AI is being used:
  • Drafting legal contracts
  • Designing rooms according to a theme and size
  • Generating written content, such as articles, blog posts, and scripts
  • Creating dialogue for characters in video games, movies, and TV shows
  • Generating song lyrics and poetry
  • Assisting with brainstorming and idea generation for advertising and marketing campaigns
  • Creating chatbot responses for customer service and virtual assistants
  • Finding the right image for a mood
  • Writing emails
  • Creating proposals
  • Writing marketing content
  • Improving emails you’ve written to be more respectful
  1. ChatGPT AI has a clear bias: the dataset that it references. The final output should always be reviewed and edited by a human to ensure it is of high quality and appropriate for the intended use.
  2. AI is going to get better with time. It uses a feedback loop to understand what humans want and what they dont.
  3. AI can improve efficiency and productivity of individuals but an also lead to job losses for some.  The market will change.
  4. Impact industries like data entry, customer service, and other tasks that rely on language. ChatGPT can be used to quickly and accurately understand and respond to customer queries, allowing companies to improve their customer service and support. This can lead to improved efficiency and cost savings for businesses.
  5. The development and deployment of ChatGPT and other language models can also create new job opportunities in fields such as data science, machine learning, and software engineering. People will need skilling.

TV Show: The cost of Internet Shutdowns

The function of a state is to protect our fundamental rights. When a government imposes an Internet Shutdown, it’s an acknowledgement that it’s unable to protect our fundamental right to free speech. Internet Shutdowns are a failure of governance. It’s also disproportionate. In order to shut down illegal speech, and sometimes they shut the Internet down when there are exams, they’re censoring all legitimate speech as well.

Digital India fails when there is no Internet. Digital Payments don’t work. Students can’t apply for exams or colleges.

India has the highest Internet Shutdowns in the world. It’s a shame.

On AI and its impact on News Media

I spoke at the Media Foundation Dialogues, organised by the Foundation for Media Professionals, held at the India International centre.

I really don’t think that journalists have anything to fear: AI tools can help them in their work, and will never entirely replace reporting. At MediaNama, we’re already using AI tools for transcription, and have started experimenting with prompts to clean copy, correct grammar, generate tweets from article, suggest article ideas, headlines, questions for interviews, speakers for conferences. We’re also creating tweets, threads, generating text to video, scripts for video. There’s so much more. The list will expand, as we become better at prompt engineering. AI will make things easier for journalists, not more difficult.

Creating and optimising videos for different platforms will be easier and faster. Processing videos will become so much more efficient.

The part that excites me most is how much translation and voice to text and text to voice will become over the next few years. That gives us the opportunity to cater to language audiences in India.

The business model challenges are of course going to increase, especially with increased advertising inventory, and fewer people required to do the same job. At the same time, this helps smaller publications punch far above their weight. How they navigate AI over the next few years is going to be critical.

A repeated some of the points I had made during the discussion, in a separate video for MediaNama: