Anna Veklich, Co-Founder of GPT4Telegram – Interview Series

Anna Veklich, Co-Founder of GPT4Telegram, is an entrepreneur and marketing strategist focused on building AI-native products and communities on Telegram. Since co-founding GPT4Telegram in 2023, she has helped grow one of the platform’s largest AI media ecosystems, operating multiple AI bots and more than ten Telegram channels that collectively reach approximately 35 million active users. Alongside her entrepreneurial work, she serves as Head of Strategic Communications & Marketing at E-Quadrat Science & Education, where she combines expertise in communications, growth, and artificial intelligence to drive technology adoption and audience engagement. Her background spans AI product development, community building, and strategic marketing, positioning her at the intersection of generative AI and messaging platforms.

GPT4Telegram is an all-in-one AI assistant for Telegram that provides access to a wide range of leading AI models for text generation, image creation, video, music, coding, translation, and research directly within the messaging app. Designed to eliminate the need to switch between multiple AI services, the bot enables users to interact with models from providers including OpenAI, Anthropic, Google, and others through a single conversational interface. The platform reports serving more than 30 million users worldwide, making it one of the largest AI experiences available on Telegram.

You co-founded GPT4Telegrambot in early 2023, before most major AI companies had established a meaningful presence on messaging platforms. What opportunity did you see at the time, and what convinced you that Telegram could become a major distribution channel for AI? 

Long before founding Chat GPT (Open AI), starting from my academic years, I always believed that AI had the potential to completely change the way we interact with information. Before launching GPT4Telegrambot, I spent over a decade at one of Europe’s leading tech universities. ITMO is the only university in the world that has won the prestigious ICPC programming championship seven times, and it is consistently recognized among global leaders in Data Science and AI. Thanks to proximity and constant collaboration with incredibly talented AI scientists and researchers, I was very lucky to observe the evolution of these technologies from the inside and understand their long-term potential. When OpenAI opened ChatGPT to the public in late 2022, it was clear something fundamental had shifted. For the first time, technology this powerful was accessible to pretty much anyone. To me, that was the moment AI stopped being a niche tool for specialists and started becoming a mass-market product. What bothered me though was the way people interact with AI. Early ChatGPT lived mostly in a web browser. But the underlying technology could easily power a chatbot, and Telegram was already one of the most developer-friendly platforms out there for building one. The barrier to entry was low and putting AI inside a messaging app just felt like the obvious next step.

GPT4Telegrambot has grown into one of the largest AI ecosystems on Telegram, serving tens of millions of users. What were the most important growth decisions that enabled that scale without relying on traditional venture capital or large paid acquisition budgets? 

Telegram was already one of the world’s largest platforms, but its potential as an AI ecosystem was still untapped. When we launched the first version of ChatGPT inside Telegram, users had very few alternatives. If someone searched for anything AI-related on the platform, our bot was often one of the first things they’d come across. Honestly, a lot of it came down to what I’d call Telegram-native SEO — smart positioning, good timing, and a product name that happened to match exactly what people were typing into the search bar. As CMO, I’m responsible for strategic marketing and growth. From the very beginning I’ve believed that in consumer AI, building a strong product just isn’t enough. You also need to understand where your users already spend their time, how their digital habits form and how to fit your product into those habits with as little friction as possible. That’s why Telegram became more than a distribution channel for us and became the natural place to scale. As a result, a large share of our early growth was organic, and we faced very little competition inside Telegram itself. It was a rare situation: barely any competition, and huge unmet demand. Today, more than 5 billion people use messaging apps. We entered the market at exactly the right moment, with a product that genuinely matched what people needed. Just 9 months after launch, we started getting acquisition offers, but we chose to keep building independently. In our first two years, we reached around 25 million unique users with almost no marketing spend. Today, the platform serves more than 36 million users and generates roughly $1 million a month in revenue and that’s all without venture funding or big paid-acquisition budgets.

Your platform brings together multiple leading AI models within a single experience. What have you learned about how users choose between models, and why is the demand for AI aggregation growing? 

The demand for AI aggregators is growing mainly because they make it so much easier for people to get started with AI. Most users never stop to think about which specific model generated an image or wrote a piece of text. What they want is simple: open an app they already know, describe what they need, and get a result right away. For them, the quality of the result and how fast they get it matter far more than what’s happening under the hood. We’re also seeing that people increasingly want everything in one place like chatting with AI, searching for information, generating images and videos, building presentations, using agentic features and all in a single platform. So aggregators are evolving from a bunch of separate tools bolted together into full AI platforms that automatically pick the best model for whatever task you throw at them. And honestly, there’s a money side to this too. Subscribe to a handful of standalone AI services on your own and your monthly bill can easily blow past $100. Aggregators give you access to multiple categories of models for a lot less, which makes advanced AI tools accessible to a much wider audience.

Having observed millions of AI interactions inside Telegram, what are the biggest behavioral differences between users who access AI through messaging apps and those who use dedicated AI applications? 

We observe a clear distinction between these two audiences. Messaging app users tend to rely on AI for everyday tasks such as translating text, recognizing images, creating holiday cards, searching for information, or generating simple content. What matters most to them is the shortest possible path to a result. They do not want to install separate applications, choose between different models, or navigate complex interfaces. Users of standalone AI platforms, on the other hand, are more likely to use AI for professional tasks. They work with code, documents, research, design projects, and other sophisticated use cases. As a result, AI embedded within messaging apps is becoming a tool for mass adoption, while standalone AI platforms continue to serve as professional-grade solutions.

Telegram recently surpassed one billion monthly active users. How has the platform evolved since you launched GPT4Telegrambot, and what advantages does it offer AI companies compared to traditional app ecosystems? 

Telegram has consistently ranked among the 5 most downloaded apps in the world for several years and remains one of the strongest tech platforms for both users and developers. Its key advantage lies in its openness to building sophisticated automated services. Developers can create bots, integrate AI models, and design end-to-end user experiences that remain far more limited across many other messaging platforms. This became even more obvious once AI agents started taking off. After AI-native Telegram products like ClaudeBot, Moltbot, and OpenClaw broke through in early 2026, you could feel developer activity in the ecosystem pick up almost overnight. Telegram itself reported that BotFather’s monthly active users jumped from roughly 3 million to 10 million in just a few months. That’s not just demand for AI bots — it’s developers actively choosing Telegram as the place to build and ship AI agents.

Through both GPT4Telegrambot and Hi, AI! Media, you have built a massive AI-focused follower base. What have you learned about audience building, community engagement, and AI content distribution that most startups still underestimate? 

Our work with our audience didn’t start with the media, it started with the product itself. It quickly became clear that most people simply didn’t understand what AI was or how it could actually help them in everyday life. That’s what led us to create Hi, AI! Media, which has since grown into one of the biggest AI-focused media networks on Telegram, with more than 15 million subscribers. What we found is that education has a direct effect on product growth. The more we talked about practical, real-world AI use cases, the faster our audience grew. If you actually want AI to become a mass-market technology, a strong product isn’t enough. Having a community gives you another strategic edge, too. It turns into an organic distribution channel, letting you test new ideas quickly and launch new products with little to no extra spend on customer acquisition.

Many users around the world either cannot access or do not regularly pay for premium AI subscriptions. What patterns have you observed in emerging markets, and how is AI adoption unfolding outside North America and Western Europe? 

In developed markets, the average consumer is already paying for 4-6 subscriptions, which creates a massive subscription fatigue. People are increasingly reluctant to add yet another recurring payment to their monthly expenses. In emerging markets, AI is often used differently. While users in developed economies tend to view AI primarily as a productivity tool, people across the CIS region, Africa, Latin America, and the Middle East are adopting it much quicker for everyday use cases from education and work to personal tasks and creative projects.

The reality  is that only about one in six people globally are actually using generative AI tools in 2025 and moreover by some estimates, over three billion people still aren’t online regularly, which tells you how wide the gap is between the most connected markets and everyone else. 

At the same time, messaging apps serve as the primary internet gateway for more than 60% of users across CIS and Southeast Asia. This is one of the reasons why we deliberately moved away from the traditional subscription model. Instead, users only pay for the features they need at a given moment. This approach works particularly well in emerging markets, where people are less likely to maintain multiple subscriptions. Overall, outside of North America and Western Europe, we are seeing a different pattern of AI adoption – practical, flexible, and use-case driven. In these markets, AI is increasingly perceived not as a standalone product, but as a universal tool embedded into everyday activities and workflows.

As major technology companies continue integrating AI into their products, how do you see the competitive landscape evolving between messaging platforms, AI model providers, and independent AI applications? 

I think the market is going to keep moving toward more specialization. Different AI companies are already carving out their own strengths and competitive edges. Some focus on search and information retrieval, some on coding and document workflows, others on visual content creation. At the same time, we’re still seeing very little competition inside Telegram itself. A few major players like Perplexity, Copilot have tried to break in, but none of them have managed to build a meaningful audience so far. About a year ago, there were talks between Elon Musk’s Grok and Telegram about a potential deal worth around $300 million, along with a revenue-sharing model for rolling Grok out to all users.It felt like a real signal that major US tech players were finally ready to go deeper into Telegram, but in the end, the deal didn’t go through. 

Over the next few years, I think competition will come down less to whose model is technically better and more to who can build a seamless user experience and actually solve people’s specific problems. As AI gets woven into more and more everyday products, distribution, accessibility, and user experience are going to matter just as much as the models themselves.

AI operating inside messaging environments raises unique questions around privacy, data usage, and platform responsibility. What ethical challenges do you believe deserve more attention as AI becomes embedded in everyday conversations? 

When AI becomes integrated into messaging environments and spaces where people discuss both personal and professional matters, the ethical concerns become significantly bigger. It is important to distinguish personal data and anonymized data, as most AI systems analyze aggregated usage patterns without identifying individual users. However, most users are not aware of how these systems work, and that is where the core challenge begins. And here’s the thing – 70% of adults say they don’t trust companies to handle their data responsibly. That’s not a small number. It tells you the industry still hasn’t done a good enough job explaining how user information actually gets handled. I believe the solution has two dimensions. The first is transparency. Users will always demand clear answers about how their data is collected, processed, and used. Companies that can provide those answers honestly and transparently will gain a significant competitive advantage. The second is digital literacy – users need to understand both the capabilities and the limitations of AI. It should be viewed as a tool that supports decision-making, not as an infallible source of truth.

Looking ahead, do you believe messaging apps will become the primary interface for consumer AI, and what developments in AI-powered communication do you expect to define in the next five years? 

I believe the next stage of AI adoption is going to be AI agents built directly into the communication channels people already use every day. Messaging apps have become a core part of our digital lives, and the natural next step is personal AI assistants that can actually take action on your behalf, not just chat with you. Telegram is uniquely positioned for that future. It’s not just a messaging platform — it’s an ecosystem for building services, media products, and automated workflows. At the same time, I don’t think the market is going to converge around one universal model. People will keep wanting choice, flexibility, and access to different capabilities. That’s exactly why AI aggregators are going to stay relevant, they let people pick the best tool or model for whatever they’re trying to do. Our long-term mission is to make AI as simple, accessible, and useful as possible. People shouldn’t have to learn how to use AI, it should just feel like a natural part of how they already work, helping them solve real problems with as little friction as possible. At the end of the day, the winners in this space won’t be the companies that build the most advanced technology. They’ll be the ones that make AI genuinely convenient, intuitive, and valuable for billions of people around the world.

Thank you for the great interview, readers who wish to learn more should visit GPT4Telegram.