Privacy-Preserving AI, Data Ownership & Blockchain Whistleblowers

Brittany Kaiser is the CEO of Alpha Tun Capital, a whistleblower and the subject of Netflix's The Great Hack.

She's also co-authored several federal bills on data ownership and AI ethics and is now building a privacy preserving AI infrastructure.

Structure for Telegram's 1 billion plus users.

So Brittany, great to have you here.

I was just telling you that, uh, before we met, I watched, uh, this Netflix documentary and was fascinated by it.

It's an area I'm very interested in.

Give us the backstory.

How did you become the subject of that documentary?

I started my career as a human rights lawyer, and when I got into technology, I realized the way that most technologies function is not uh really congruent with the way that human and civil rights actually function.

So, technology needs privacy preserving features, it needs data protection in order to make sure that Individuals, children especially, as well as companies, can use it and make sure that their most personal data sets and other sensitive information are not shared with third parties without their consent, right?

And even maybe make money off of that data, which is another thing that some people have been interested in as well.

Yeah, absolutely.

And I think when we, when we start to implement privacy centric technologies, which has been my concentration for the past decade, we have the opportunity to both protect our rights and also build very successful companies that are highly ethical and compliant.

OK, so let's talk about your company, Alpha To Capital.

What does it do?

What's your mission?

Alphaton Capital is focused on building privacy preserving technologies and making them accessible to the entire world.

We've been investing in one of the world's largest privacy ecosystems, which is the Telegram Super app.

It's one of the largest, most used apps in the entire world and launching both easy to use applications for business, gaming, decentralized finance, as well as large scale investing in confidential compute, which is actually the hardware.

That goes in data centers that allows everybody around the world to use AI and other data processing.

From the core, we are focused on having a fully vertically integrated ability for people to fully preserve their privacy while using any type of technology, especially in the age of AI.

Yeah, now why Telegram?

Of course it's huge and it's global.

Is there anything else about it that makes it particularly conducive to what you're doing and, and what's your ecosystem there?

It's the only super app that's used by over a billion people that has been built with data protection and privacy features as the core.

It also has fully vertically integrated blockchain technology, so you can launch Web 3 technologies also in this app.

There's no other daily used app by billions of people that allows you to do that, and I think a lot of companies in the digital asset space have tried to launch. iOS and Android apps, and have either been rejected or had trouble actually launching their business there.

Or they're giving away 15 to 30% of their revenue, where in Telegram, it's a permissionless open source app store, so you can actually just launch your app, even with web 3 features, where you are using tokens on the back end and launch it to a billion people.

There's, it's an unrivaled opportunity.

So Cocoon AI for everyone, first explain what is Cocoon AI and talk about that concept and why that's important.

I think the easiest way for people to digest Cocoon AI is firstly to understand that it is the XAI to X is Cocoon AI to Telegram.

So Telegram is nearly twice the size, uh, user-wise as as the X platform, and it is now fully vertically integrating its own AI.

It's the first privacy-centric.

And decentralized AI inference network.

So that means when you are processing AI, all of the data that you are putting in and all of the intelligence that you're getting out is leaving all of your data with you.

It's not sharing it with third parties.

There is no other widely used AI network that works that way.

So Kakun AI is you needs telegram.

I did not know that.

Very interesting.

Um, so let's talk about data sovereignty.

So now there's growing concern about intellectual property and the platforms like OpenAI that may not be fully owned by the people who built the IP.

So how does Alphao's approach differ?

Our approach is making sure that all of your data is yours, whether you are an individual, a company, a government department, an organization, if you are running AI.

You want to make sure that all of your IP, your copyright, your material non-public information, if you're a public company like we are, or your classified information if you're a government, remains yours and remains on your devices.

When you are running most AI today, basically every platform that everyone has access to, you are giving your data to the owners of the AI platforms and sometimes millions of other databases around the world without your explicit consent, because you probably didn't read the terms and conditions.

Who does, right?

I mean, come on.

But you know, it's very important that we're familiar with that.

Now.

You've deployed hundreds of Nvidia's most powerful GPUs, um, real capital, real hardware, um.

Who pays for that compute and how does that turn into revenue for shareholders?

Yeah, so, uh, when we buy and install GPUs in our data centers, we provide both regular and confidential computing to our clients.

Our clients are either licensing by the hour, by the month, or multiple years in advance to get access to our compute power.

Usually that's because they're running enterprise grade AI deployments, and so they want to lock in that they actually.

Have that compute power for the foreseeable future.

So they pay us to use our GPUs and shareholders are able to take advantage of those very steady and very low risk returns.

Yeah, now let's talk about Agentic AI because that's been exploding lately.

You know, they'll make dinner reservations for you, even call customer service, things like that.

So how does Alpha 10 fit into the agentic AI picture and why does privacy matter when it comes to agentic AI?

So most people that are using AI agents today are hooking it up to all of their most sensitive data sets, and what that means is everything about your life is now accessible to this piece of software and probably the owners of that software and anyone who is good enough to to write a hacking prompt.

And so that is rather dangerous and uh creates a lot of cybersecurity issues that weren't there before.

So, last month when we were at the World Economic Forum in Davos, we announced that we had started launching a 100% privacy preserving AI agents, so that you own all of your data the entire time, fully vertically integrated on confidential computing.

So any data that you are giving to the agent just stays with you on your machine and is not leaving the third party database.

Now the Government Blockchain Association just released a major report on using blockchain to fight fraud, waste, and abuse, and you were invited to speak at Capitol Hill about this.

So how does Alphaun fit into that mission?

So, as a former whistleblower, I know that really the definition of that legally is to provide evidence of fraud, waste, abuse, and other criminal activity.

And so we have been looking at privacy preserving ways to help whistleblowers report evidence that they have.

As you've seen, everyone from our president to Scott Besscent to Elon Musk over the past many months have been encouraging whistleblowers to report what they've seen so that government agencies can actually take these companies to court and win against these fraud cases and then give a piece of that back to the whistleblowers.

It's usually 15 to 30%.

So we announced yesterday that we launched an application called Vera Report.

Vera allows whistleblowers to securely and anonym.

Anonymously submit information, get it verified, open up a case, and then KYC in order to receive their whistleblower award.

And they submit that on the blockchain, so that keeps it immutable and also a private.

You also have some serious names on your table.

So Michael Turpin, others board members from Stanford and the NASDAQ, are institutions starting to pay attention to Alphatan and talk about like just the conversation that you're having with hedge funds, family offices, institutional investors.

I think, uh, over the past, uh, 5 months since we became public, uh, since we've been touching digital assets, a lot of the, uh, a lot of the public markets have started to try to understand tokens on a balance sheet, and they've really had trouble valuing that.

And so now that we have shored up our balance sheet with hardware.

With Nvidia's latest GPUs, which are some of the most well understood investments in the entire public market, and we have just started to deploy clusters at an enterprise grade level.

We're getting a lot of excitement from funds, family offices, and individual investors all around the world who are waiting to see our first big results that are coming out.

We're very excited about these huge deployments because we're starting to get to hyperscalar level in terms of the equipment that we have access to and the compute power that we're able to provide.

OK, Rennie, fascinating area that you're in and curious to hear how all this goes for you.

Thank you so much for coming.

Absolutely, thank you so much.