The Blockchain and AI Investments and the Hopes for Changes in Innovations’ Regulation

For years, the discussion around regulating blockchain and cryptocurrencies has been as polarized as it is pressing. On one side, proponents of decentralization argue that regulation undermines the very foundation of these innovations. On the other, there’s a growing recognition that clear oversight is essential.

During Donald Trump’s new presidency, the crypto world hoped his administration would establish a clear path forward, perhaps setting a global precedent. Yet, despite some steps, there are more milestones ahead. This lack of clarity creates disparities in investment in blockchain, AI, and digital assets, harming the interests of key stakeholders.

The Need for Regulation in Blockchain and Cryptocurrencies

Why does it matter? Blockchain, AI, and digital assets are deeply intertwined. Without clear rules, we will face the bumps in the road one of which all witnessed on 27, January, 2025 during the stock rout in the global AI sector.

We saw how Trump canceled all key executive orders issued by the 46th U.S. President, Joe Biden, that limited the circulation of crypto in the U.S. However, the new order prohibits the creation and use of central bank digital currencies (CBDCs) in the U.S..

Despite potential discussions about the pros and cons of a digital dollar, it cannot be denied that we will see a rise of CBDCs in the near future. One key economically significant country in the Middle East, the UAE, is actively working on the concept of a digital dirham. The leader of Greater Asia, China, has an interest in scaling up the digital yuan. The U.S. will have to engage in global CBDC competition, whether they want to or not, and private dollar stablecoins represent competition on a completely different corporate level.

We also saw that the SEC changed its interpretation of existing legislation, softening its stance on cryptocurrencies. The SEC rescinded rule SAB 121—now banks and companies are legally allowed to store customer cryptocurrency.

Regulatory Progress and Remaining Challenges

However, there remains a need to enshrine changes in federal legislation analogous to the Securities Act. It is no coincidence that Trump established a working group on digital assets through the National Economic Council. This group includes representatives from the Treasury Department, Justice Department, SEC, and CFTC. The working group is led by Trump’s advisor on AI and cryptocurrencies, David Sacks, former COO of PayPal. Elon Musk also briefly played a role in PayPal as CEO.

So far, there is no 100% green light for U.S. banks to hold, sell, or buy cryptocurrencies directly; almost nobody will risk direct exposure to cryptocurrencies while waiting for fundamental changes in federal laws. Of course, U.S. banks will accelerate their practices to operate with Bitcoin, Ethereum, and other possible crypto ETFs; they will work with crypto indirectly via separate legal entities and heavily invest in blockchain, AI, and crypto sectors while placing great hope on AI-driven financial services markets and tokenization of real-world assets as promising aspects of realization.

However, investments will not be fully freed up. Moreover, the legal status of AI autonomous assistants demands no less clarity than the crypto issue itself. AI can earn money but cannot yet be legally considered a person. It is necessary to create DAOs (Decentralized Autonomous Organizations) so that AI can manage funds, hire people, and purchase GPUs.

The main problem is that AI cannot be smarter than humans in textual form; it makes similar mistakes as humans. AI can surpass humans in speed and solving specific tasks but overall remains below humans since it does not receive information from human brains managing businesses or people.

Even if most or soon nearly all repetitive operational processes in businesses are automated using AI but 1) unique processes will require interaction between humans and AI; 2) strategic decisions remain with humans. In both cases (1) and (2), AI will assist but not in everything; where it helps—errors occur.

Networks of AI Assistants (AI Agents)

Autonomous AIs raise questions about their regulation; without recognizing blockchain transactions as equivalent to paper transactions everything will stall. AIs must have choices; they will determine whether certain cryptocurrencies or tokens are needed and whether they provide benefits. LLMs are not designed to be knowledge bases. You should never expect accurate information from text generated by them; therefore keeping humans in the loop is essential to catch errors made by AI.

Still, AI has one fundamental problem: since it is not human, all human pains are alien to it—principally incomprehensible. Good design (for example, a website) always involves empathy (if we are talking about a site for people rather than AI systems).

There is no anonymity in blockchain, which means there is accountability, even if AI is managed through a DAO. The emergence of autonomous AI agents is indeed reshaping the landscape of human-AI interaction, leading to a situation where the need for human prompting may diminish. These agents are designed to operate independently, engaging in proactive decision-making and executing complex tasks without requiring constant human input.

Regulation is also needed to pave the way for further evolution of large language models (LLMs). As of January 20, 2025, the Chinese DeepSeek cost approximately $5.5 million, compared to billions spent on key LLMs developed in the U.S. in total.

What’s the clue? Of course, China has an experience on how to effectively use blockchain while the People’s Bank of China has been evolving digital yuan for the last years.

Another point. The open-source nature of many LLMs allows for the development of new LLMs at a lower cost. This presents a crossroads: we could either see the rise of Superintelligent Artificial Intelligence (SAI) based on open-source LLMs, making SAI effectively cheap and accessible to the public, or AI evolution could take another direction. New iterations of AI programs may not be available to the public.

Regulation must figure out how to encourage creators of LLMs to progress by attracting investments into their projects while creating mechanisms to protect the rights of both creators and investors. Otherwise, valuable breakthroughs in AI developed within companies may remain inaccessible to the public, as these companies fear that other players could exploit their advancements to create more advanced LLMs.

For example, DeepSeek's success highlights the effectiveness of collaboration between human and digital labor to achieve remarkable results with lower costs. Yes, they enhance the experts system presented already at many LLMs but the DeepSeek R1 and Janus-Pro-7B can put on by themselves the role of some expert and this way functioning minimizes the energy consumption.

The novelty is how DeepSeek learned to read faster than many other LLMs catching the whole phrases instead of going from word to word. At the same time, the AI data-centers and relevant expenses on them, including chips, don’t disappear. Energy is also a big factor in this equation.

This example resonates with me as I served as Chairman of the Board at “Otkritie Broker” from 2017 to 2022. In 2022, our share of in-house development of innovative tech systems within the “Otkritie” ecosystem reached 70%, marking a significant milestone as we achieved a digital breakthrough primarily using our corporate resources. We introduced a new standard of digital communications in the industry and did this mostly with internal corporate resources.

The Intersection of Blockchain and AI in Fintech

The case of DeepSeek caused the shock on Wall Street on 27, January, 2025, when key US and global AI and AI-chips related companies lost double digit in one day of its capitalization. This only shows how it’s very important not only to attract investments into AI (the U.S. companies did that) but to have sustainable communications with investors to constantly explain what’s going on in your business.

LLM isn't the final destination of an AI journey. Having the best LLM isn’t the goal of US big tech companies, they want more. AI-Robots, AI-driven autonomous cars, AI-assistants and cryptocurrencies economy for them. There’s much bigger fish than having the best LLM. And if this point could be properly presented to investors then we would never see the phenomenal dip of the U.S. stock market when for one day more than $1.2 trillion have been wiped out.

Another regulatory topic is how to protect business models from being copied. In an internet world where "internet goods" can sometimes be viewed as "free gifts," value increases as these "gifts" attract more attention from users. The competition for attention has emerged as a new currency.

AI also demands a transformation for the Internet. Websites and social networks will need agent-friendly versions for autonomous AI assistants. We are approaching a time when there will be an Internet for humans and an Internet for digital AI-driven entities, leading to a convergence that creates space for physical AI proliferation.

However, success will not hinge on who can prompt AI better; it will depend on who understands their customers better. The downsides of AI include a lack of understanding of the physical world, the lack of emotions and empathy memory, and incapacity for strategic planning.

Regulation must not lag behind innovations but rather stimulate their creation and investment. This applies to blockchain implementation as well; blockchain technology could enhance transaction speeds and efficiencies — South Africa's Reserve Bank has significantly reduced settlement times using blockchain solutions. And this fact was vastly discussed during the World Economic Forum-2025 in Davos.

The main obstacle to widespread blockchain adoption is the lack of clear statements from regulators affirming that deals made on the blockchain hold the same weight as paper contracts signed by parties involved.

Investors get a gut feeling that the future lies in blockchain transactions; hence Apple had an unsuccessful start to its year—shares fell about 9% in the first three weeks of the year, while Apple Pay remains popular.

Overall, we see how blockchain and AI serve as backbones for modern fintech with potential to drastically lower expenses for players in the financial sphere. Regulation in all countries must fully unleash this power.