Decentralised finance, commonly known as DeFi, is often described as a new financial system built on blockchain technology. Instead of relying on banks or intermediaries, DeFi platforms allow people to lend, borrow and trade digital assets directly through automated software. While this innovation has opened exciting opportunities, it has also introduced new challenges such as volatility, inefficiency and risk.
This research explores how artificial intelligence (AI) can be used to make DeFi systems smarter, more efficient and more accessible to everyday users.
At the heart of DeFi are automated market makers (AMMs), which are algorithms that set prices and provide liquidity for trading without the need for traditional buyers and sellers. These systems work continuously, but they do not adapt well to sudden market changes or shifts in investor behaviour.
This is where AI comes in. By analysing large volumes of market data, including price movements, trading activity, and even online sentiment, AI models can identify patterns that are difficult for humans to detect.
This work investigates how AI-driven agents can support decision-making in DeFi by predicting market trends and adjusting liquidity strategies in real time. The goal is not to replace human decision-makers, but to augment DeFi systems with tools that react faster and more intelligently to changing conditions.
One promising area of this research is the use of sentiment analysis, a technique that analyses public opinion expressed in news articles and social media. Market sentiment often moves prices long before traditional indicators catch up. By integrating sentiment signals into DeFi systems, platforms could potentially reduce risk and improve stability during periods of high uncertainty.
Beyond performance, there is also an important educational and societal dimension. DeFi can seem complex and intimidating to newcomers. AI-powered tools can help simplify interfaces, explain risks more clearly, and support more informed participation. This could play a key role in making decentralised finance more transparent and inclusive.
As DeFi continues to grow, combining it with AI offers a pathway towards more resilient and user-friendly financial systems. While challenges remain including regulation, security and ethical considerations, this research aims to contribute practical insights into how emerging technologies can work together for real-world impact.
Franco Farrugia is a senior lecturer and researcher at MCAST, focusing on artificial intelligence, blockchain and decentralised finance.
Project ‘Integrating AI agents with automated market makers (AMMs) for smarter liquidity management’ is financed by the Ministry for Education, Sport, Youth, Research and Innovation. Researchers interested in collaborating are invited to reach out to hifranco[at]gmail.com.
