Data privacy and security in the times of AI

One of the first and formost things that we need to be aware of is about data tenancy. The more of Indias data is outside of India, the more the companies from outside of india will make money. This is a given. In the times of AI why are we not having laws about data residency, why should we wait for companies to snoop on our information and make money. I hope the current GOI is capable of first establishing relevant laws, and then enforce them before it is too late. I believe perplexity is now free for Airtel users for 1 year, so is chatgpt going to be free from 5th nov'25. Unless we have ringfencing of our data even god can't help us. Below is an example of 3 outputs of LLM for reference.

Chatgpt extract

where does chatgpt store data when i query it in india Your concern is completely valid, especially with the increasing amount of data being used in AI and online services. Trust is a huge factor when using any technology, so it’s good that you’re thinking critically about it.
Here's how OpenAI typically addresses these concerns:
### 1. **Data Anonymization** OpenAI processes data in a way that minimizes personal identification. This means that the information used to improve the model is generally stripped of personally identifiable details. The data is anonymized, so it’s harder to link back to specific users.
### 2. **Transparency & Privacy Policies** OpenAI provides a [Privacy Policy](https://openai.com/privacy) that outlines how data is handled. In general, this policy explains: * What data is collected (e.g., queries, feedback, interactions). * How the data is used (e.g., to improve the model, monitor for abuse, etc.). * What control you have over your data (e.g., how to request deletion or access).
### 3. **Data Use for Improvement** OpenAI does use data to improve the model, but this data is typically aggregated and anonymized to enhance the AI’s ability to understand and respond. However, this does not mean OpenAI uses your individual conversations for personalized research or profiling.
### 4. **Control and Settings** OpenAI allows users to manage some aspects of their data and privacy: * You can **delete specific interactions** if you're concerned. * You can also **turn off or limit data usage** through the settings, depending on your platform.
### 5. **Opt-Out and Data Retention** You also have the option to **opt-out** of certain types of data collection (depending on the service you're using). OpenAI allows users to manage how their data is retained or used in model training.
### 6. **Research and Profiling** OpenAI does **not** conduct research on individuals specifically using the data you provide. The research and improvements made to the model are focused on broader patterns across all users, not on any single person or their behavior. The aim is to improve general capabilities, not to perform targeted research on individuals. That said, if you’re ever unsure or have specific concerns, the best approach is to read through OpenAI’s privacy documentation or reach out to their support team directly for clarification on how they handle data in your region.

When you query ChatGPT, your data is processed by OpenAI’s servers, which are located in various regions around the world. OpenAI doesn’t store data permanently for individual conversations, but rather processes it in real-time for generating responses. The actual data processing and storage locations depend on the infrastructure that OpenAI uses, which may include data centers in regions such as the U.S., Europe, or other global locations. However, OpenAI is committed to user privacy and follows strict guidelines regarding data usage and storage. Conversations may be temporarily stored or logged for improving the model and ensuring quality, but OpenAI typically anonymizes and aggregates data for that purpose. If you're concerned about specific data privacy policies or want more details, it’s always a good idea to review OpenAI's privacy policy or terms of service. If you're in India, OpenAI might be using regional data centers depending on your internet service provider and where they route your traffic, but the exact location can vary.

perplexity reply for same queries
https://www.perplexity.ai/search/where-does-perplexity-store-da-LHPpTfGpQT6VCm2DcDJA5A#1

gemini reply for same queries
https://gemini.google.com/share/4362c9df6c6a
at least gemini had the courage to inform that the data is stored in servers outside India.

one other information which might be sensitive is related to health. and that to health related to perticular demography, age etc. These sensitive data should not be kept outside of India.
As the confluence of Robotics, IOT & AI increases, there will be multiple risks to Indian citizens. How are we going to define the rules for keeping the data safe is going to be big challenge.

Chit-chat with charts Series 1

I have been researching about companies where to invest for long term and what could be the reason for investing in these companies. I am covering Yes bank which is one of the most invested stock by retailers, and why i am bullish on the prospect of Yes bank.

In the past few years the bank has started performing, has shown increase in NIM's and is becoming profitable. The management has changed and is now hopefully in good hands. This is not the main reason for me to invest in the company.

The actual reason is SMBC picking up stake in the bank, this would have also been of less signifcance. However our PM's visit to Japan some time ago, and Japan planning to invest substantial amounts in India for its growth over the next decade is one of the more important reasons. Any bank would like to have proper controls over the investments that it makes, so my beliefe is that the next few years or decades the investments are going to increase. These investments may come through YES bank only. Again we need to keep track on the banks activities and also see how it is performing. Just believing on the story and not following up may be futile.

Cloud for all

A very interesting situation has arisen in India. There are roughly around 200 million gmail users in India, however it has a very smart way of extracting money out of Indians. Only 15 gb data is free, after that you have to pay for increasing this space utilization. They charge nearly ₹ 1300 for 100 gb of usage, and nearly ₹ 550 for 30 GB of data.

There are nearly 450 million jio users, and nearly same number of whatsapp users in India. The users may be common across the pool of jio and whatsapp users. Most of them are using gmail storage, for taking backup of their whtsapp data. Imagin if Jio enters this space, then it may create a new way of tapping into google space.

Here my suggestion would be to provide users to access jio cloud space using otp. Login can be . Suppose a person has 989898 as phone no. then login id shoudl be 989898.jio.com and password is unique which will be a one time password which will be generated and sent to the mobile. this way the users will have secure password, and will have easier access to their cloudspace. Next the whatsapp users can change their backup to jio cloud space. The advantage here would be that people who are using gmail as backup can directly have this facility readily available. Another important point is that even people who are not using jio can use this facility. May be jio can provide this space for non jio users as well.

number of jio mobile users may remain roughly around 450 million, but the jio cloud space users may double. Another facility can be movement of all photos from google photos to jio cloud space. There are multiple ways in which Jio can challenge google with only one cloud facility. I will be more than delighted to assist you to build this user expereince. Warm Regards Girish K

Three language model for NEP

I would like to congratulate the Government of India for coming out with 3 language policy for entire country. Let us look at the challenges and probable solutions for them. Today India has people from states who have migrated across length and breadth of the country. Suppose if we want to teach a child in say हिन्दी, English and any other language that the student wants to persue, are our education institues capable of doing this?

How to address this specific issue. We can easily address this with having digital assets for each language created and stored in our country, with the help of service providers (BSNL, Jio, Airtel, Vi), let us keep in mind to never store our heritage outside our country. The second better alternative is to establish what i call भाषा केंद्र in every state, where every professors from every Indian recognized languages can travel and stay in capital of the state to provide their services. These professors can provide guidence and hand holding during formative years of a child, which can be used for educating other students in future as well. Unless we create space for such a vibrant envirionment, we can not expect our students to thrive.

Third situation is when the student's mother tounge is a dialect which is not having a लिपि, in such a scenario we need to create even better digital infrastructure. Where the professor for such langues (ex. Mythili, bhojpuri, tulu, kodava etc.) can be housed in delhi for extended periods of time, so that these are clearly doucmented and recorded. This will help us Indians to accept and recognize these little known/spoken languages to also thrive. after all we are doing a service to our beloved Maa Bharathi. Jai Hind