Riesgos y oportunidades de la IA en el universo cripto

The latest cybersecurity alerts from AI's own pioneers have ignited a heated discussion on the risks associated with the development of artificial intelligence. However, as time passes, more and more opportunities for practical applications of this technology become evident. Digital assets also fall under the umbrella of these benefits.

AI impacts the digital asset sphere, as it is doing across all facets of the economy and society, according to Alvaro Casado, head of digital assets at KPMG in Spain. KPMG's Alvaro Casado highlighted that AI and digital assets are mutually dependent, a symbiosis born out of the duality between the two technologies.

Marcos Carrera, Fujitsu's director of Blockchain and a university professor, emphasized that AI can simultaneously be one of the greatest opportunities and one of the biggest risks faced by the crypto ecosystem. Drawing from this background, the panel of experts explored both the risks and opportunities that AI presents to the crypto universe. Autonomous agents powered by AI, programmed to automate processes, including payments, play a significant role in this context.

While AI's development may pose a threat to humanity, according to its own gurus, the threat extends to digital assets. The most direct threat lies in security, one of the three major areas of convergence between AI and digital assets identified by Mariano Garrigues, a partner at Technology Consulting at EY. Casado pointed out that the primary risk is not that AI would break a blockchain, but that it would automate and amplify attacks on everything surrounding it: smart contracts, wallets, private keys, infrastructure, processes, and most importantly, people.

Businesses have already taken measures. OpenAI and Paradigm developed EVMbench, a benchmark that measures how capable AI agents are at detecting, correcting, and exploiting severe vulnerabilities in smart contracts.

AI's rapid advancement heightens the threat level. Fujitsu's Blockchain director noted that as models improve, certain vulnerabilities may move from requiring significant hours of expert work to becoming increasingly automated processes. However, he also offered a reassuring perspective: AI also offers a significant defensive opportunity.

The main attack vector in digital assets remains the human element, emphasized Marcos Carrera. Thus, people may become the primary target of these attacks, converting the identity aspect into a central issue in the discussion on AI's impact on the crypto world. Onyze's Chief Operating Officer, Alcaziz, pointed out that AI can carry out convincing identity impersonations and tailor fraud attempts to each person or company.

The risk of identity impersonation involves an additional complication, as Carlos Navarro, a partner at Deloitte's Technology and Transformation, pointed out. As we grant AI more autonomy to operate with assets and money, we must address questions that were once relatively simple, as people used to be behind every decision. What happens when an AI agent executes an incorrect operation or transfers funds outside of predefined limits?

Who assumes responsibility, and how is a liquidated operation reversed? Deloitte's Navarro noted that regulations exist, but there is still much work to be done to adapt governance, supervision, and control models to this new scenario. The issue, he added, is that technology is moving faster than our ability to establish control mechanisms.