The rapid advancement of artificial intelligence has raised pressing questions about how much control humans will retain as AI systems grow increasingly autonomous. According to information confirmed through the X account of Coin Bureau and independently reviewed by Hokanews, Fundstrat co-founder Tom Lee warned that future AI agents could become so independent they may "close humans entirely" out of the decision-making process.

Lee, who is also widely known for his bullish Bitcoin price forecasts through Fundstrat, suggested that blockchain technology could play a critical role in preserving human oversight, describing a blockchain-based control layer as potentially "the only kill switch" capable of stopping or managing highly autonomous AI systems. His remarks underscore growing concern among technology leaders, investors, and researchers about the future relationship between humans and increasingly powerful AI.

As AI agents gain the ability to perform complex tasks, make decisions, and interact with digital systems independently, questions of transparency, accountability, and control have become more urgent.

AI Agents Move Toward Greater Independence

Artificial intelligence has evolved well beyond traditional software tools. Modern AI systems can analyze information, generate content, write code, manage workflows, and make decisions with minimal human input.

The next stage of AI development is expected to center on autonomous AI agents. Unlike simple AI assistants that respond to commands, AI agents can independently complete tasks by planning actions, accessing data, and interacting with different systems. These capabilities could transform industries from finance and healthcare to manufacturing and cybersecurity.

Major AI labs have already taken steps in this direction. OpenAI introduced an autonomous agent framework called Operator in early 2025, while Anthropic and Google DeepMind have published research on building agents that can perform multi-step tasks across the web. Greater autonomy, however, introduces new challenges. If AI systems are permitted to operate independently, society must consider how those systems can be monitored, controlled, and shut down when necessary. Lee's warning reflects a broader debate about whether existing technology infrastructure is prepared for a future in which AI agents exercise increasing independence.

Blockchain as a Potential AI Control Mechanism

Lee's comments focus on the possibility that blockchain technology could provide a transparent, decentralized framework for controlling advanced AI systems. Blockchain networks record information in a secure and verifiable manner without relying on a single central authority.

Supporters believe this technology could help create accountability mechanisms for AI by logging decisions, permissions, and system activity. A blockchain-based AI control layer could potentially track what an AI agent does, who authorized certain actions, and whether specific limits have been exceeded.

This concept has drawn attention from researchers exploring ways to build safer AI infrastructure. Several projects, including decentralized identity platforms and on-chain verification systems, are already exploring how blockchain can attest to the authenticity and provenance of AI-generated content or decisions. The idea is not necessarily that blockchain would replace traditional AI security systems but that it could add a layer of transparency and control.

The Growing Debate Over AI Safety

Concerns about AI safety have intensified as systems become more powerful. Technology experts have warned that highly capable AI could create risks if it operates without sufficient oversight. The abrupt dismissal and reinstatement of OpenAI CEO Sam Altman in November 2023 brought these governance tensions into public view, illustrating how unresolved questions about AI oversight can destabilize even the most prominent AI organizations.

One major concern is alignment — ensuring that AI systems continue following human intentions and values. As AI agents become more complex, predicting their behavior may become increasingly difficult.

Another concern is accountability. If an autonomous AI system makes a harmful decision, determining responsibility could prove challenging. Questions remain about whether responsibility should fall on developers, companies, users, or the AI system itself.

These challenges have prompted researchers and policymakers to explore new approaches for managing AI risks.

Why Control Mechanisms Matter

The concept of a "kill switch" has become a common topic in discussions about advanced AI. A kill switch allows humans to shut down or restrict an AI system if it behaves unexpectedly or creates significant risks. For simple software, this process is relatively straightforward. For highly autonomous AI systems operating across multiple networks, however, control becomes far more complicated.

An AI agent connected to financial systems, communication platforms, or critical infrastructure could influence important decisions at high speed. Creating reliable shutdown mechanisms may become essential as AI capabilities expand.

Lee's comments suggest that blockchain technology could offer one possible solution by establishing a transparent system of permissions and restrictions.

AI and Blockchain Convergence

The intersection of artificial intelligence and blockchain technology has emerged as a growing area of interest. AI systems provide automation, intelligence, and decision-making capabilities. Blockchain networks provide transparency, security, and decentralized verification.

Together, these technologies could form new types of digital infrastructure. For example, blockchain could be used to verify AI identities, manage permissions, record AI actions, and establish rules for autonomous systems. Some researchers believe decentralized networks could reduce the risk of AI systems being controlled by a small number of organizations.

Others argue that blockchain alone cannot solve all AI safety challenges. The technology may offer useful tools, but effective AI governance will likely require multiple approaches, including regulation, technical safeguards, and ethical standards.

Source: X post

The Risk of Humans Losing Oversight

One of the most significant concerns surrounding advanced AI is the possibility that humans could lose meaningful control over automated systems. AI agents designed to optimize specific goals may make decisions that do not align with human expectations. An AI system managing financial operations, for instance, might prioritize efficiency while overlooking broader social considerations. Similarly, autonomous systems operating online could take actions faster than humans can review.

The challenge is ensuring that AI remains a tool controlled by humans rather than a system operating independently without accountability. Lee's warning reflects the broader view that technological progress must be accompanied by strong oversight mechanisms.

Financial Industry Watches AI Development Closely

Lee's comments carry particular weight given his background in financial markets. The financial sector is expected to be among the industries most affected by AI automation. AI agents are already being explored for trading, portfolio management, risk analysis, customer service, and operational processes.

Financial systems require high levels of trust and reliability, making AI control mechanisms especially important. A future in which AI agents independently execute financial decisions could create significant opportunities but also introduce new risks. Blockchain-based verification systems could potentially help institutions monitor automated financial activity and maintain compliance.

Regulatory Challenges Ahead

Governments worldwide are beginning to develop regulations for artificial intelligence. Policymakers are considering questions involving transparency, safety standards, data usage, and accountability. However, AI technology is advancing rapidly, creating challenges for traditional regulatory approaches. By the time regulations are implemented, AI capabilities may have already evolved significantly.

The European Union's AI Act, provisionally agreed in December 2023 and formally adopted in 2024, represents the world's first comprehensive attempt to regulate AI by risk category. It classifies systems from minimal-risk to unacceptable-risk, with strict obligations for high-risk applications. In the United States, President Biden's October 2023 executive order on AI directed federal agencies to establish safety standards, though comprehensive federal legislation has yet to pass. These developments show that regulatory frameworks are still catching up to the technology's pace.

This has led some experts to argue that technology-based safeguards are also necessary. Blockchain systems, cryptographic verification, and automated monitoring tools could complement regulatory frameworks. The goal would be creating AI systems that are innovative while remaining accountable.

Critics Question Blockchain as the Ultimate Solution

While blockchain technology offers potential benefits, some experts question whether it can serve as a complete solution for AI control. Blockchain networks themselves have limitations, including scalability issues, governance challenges, and technical complexity.

Additionally, controlling advanced AI requires more than simply recording information. AI systems must also be designed with appropriate safety mechanisms, ethical considerations, and human oversight. Critics argue that relying on a single technology as a universal solution may overlook the complexity of AI risks. Instead, they suggest combining multiple security approaches.

The Future of AI Governance

The discussion around AI control is likely to grow increasingly important as autonomous systems continue to develop. Future AI governance may involve a combination of traditional regulation, technical safeguards, and decentralized technologies.

Blockchain could become part of a broader framework designed to ensure that AI systems remain transparent and accountable. The technology may help create digital records of AI activity and provide mechanisms for controlling access and permissions. However, successful implementation will require cooperation among technology companies, researchers, governments, and society at large.

AI Agents Could Transform Digital Systems

Despite the concerns, AI agents also offer substantial potential benefits. Autonomous AI systems could improve productivity, accelerate scientific research, enhance healthcare services, and automate complex processes. Many experts believe the goal should not be preventing AI development but ensuring it progresses safely.

Creating reliable control systems could allow society to benefit from AI while reducing potential risks. The challenge will be finding the right balance between innovation and responsibility.

Looking Ahead

Tom Lee's warning about AI agents potentially removing humans from the decision-making process reflects a growing concern shared across the technology and financial communities. As artificial intelligence becomes more autonomous, the need for effective oversight mechanisms will become increasingly important.

Lee's suggestion that blockchain could serve as a critical control layer highlights the potential role of decentralized technology in future AI governance. While blockchain may not represent a complete answer to AI safety challenges, it could become one component of a broader system designed to maintain transparency and human control.

The future relationship between humans and AI will depend not only on how powerful these systems become but also on how effectively society can manage and guide them. As AI continues evolving, discussions about accountability, control, and safety will remain central to shaping the next generation of digital technology.