The convergence of AI and Distributed Ledger Technology

The convergence of AI and Distributed Ledger Technology – opportunities and risks

Artificial intelligence (AI) and Distributed Ledger Technology (DLT) are transforming industries. Each of these technologies can offer benefits, but by combining the predictive and analytical capabilities of AI with the decentralised and resilient infrastructure of DLT, this synergy of "AIxDLT" has the potential to accelerate the transformation of a range of products, services and industries by enabling smarter, automated systems that operate with improved transparency, efficiency and trust.

In this paper, developed by Deutsche Bank and Clifford Chance, we have brought together our resources and insights from our financial services and legal perspectives to further the conversation about AIxDLT, highlighting key practical considerations for its implementation and some of the legal, commercial and technical challenges that must be navigated.

Successful implementation of AIxDLT – key takeaways

Successful and responsible exploration of AIxDLT requires internal and external engagement, a holistic approach and significant investment in infrastructure and talent to fully realise the potential of these technologies. In this paper, we share five practical insights for organisations developing or considering exploring AIxDLT use cases:

Take a holistic and collaborative approach to use case exploration and implementation: Organisations must consider the impact of proposed projects and business models on their daily operations and adopt a holistic and collaborative approach to implementation. This involves understanding the implications for the entire business and fostering collaboration across various departments, including legal, technology, risk, compliance, sales, HR, finance, and tax.
Build internal capabilities and technology literacy to support the changes ahead: Building internal capabilities and literacy in AI and technology is crucial to leverage these opportunities fully. Staff and stakeholders need to understand the capabilities and limitations of the technologies being used to ensure that use is appropriate and that there are checks and guardrails.
Consider wider impact and alignment with firm policies and culture: Organisations must also align new technologies with their policies and culture, considering factors such as customer care, transparency, accountability, energy consumption and sustainability goals.
Strategically assess and navigate legal obligations and risk: Navigating the complex legal and regulatory landscape requires early due diligence, strategic risk management (including through the contractual framework) and collaboration with trusted advisers.
Consider engagement with policymakers, regulators and industry stakeholders, and anticipate new legislation: Engaging with policymakers, regulators and industry stakeholders can also be crucial to staying ahead of regulatory changes, understanding sectoral norms and regulatory expectations and shaping emerging policies.

AIxDLT – use cases

There are a range of use cases emerging or being explored in the finance sector with differing levels of maturity. Imagine a world where a fully autonomous agent could use a person's digital identity and preferences to book and pay for a weekend in Paris across different service providers, or a supply chain system where AI analyses vast troves of data to predict liquidity needs and then automatically triggers smart contracts to optimise working capital. Some AIxDLT use cases are already being tested or rolled out, for example in areas such as treasury management where platforms are being tested and launched to support 24/7 global treasury operations through AI-powered systems optimising cashflows across multiple currencies and jurisdictions in real time and use of DLT to ensure that all transactions are recorded securely and immutably.

In this paper we have focused on four use cases for AIxDLT that have the potential to transform a range of industries:

Improvements to smart contract development and performance through the ability to test and interrogate their functionality.
AI-powered blockchain oracles for enhanced reliability when connecting distributed ledgers to real-world data.
Controlled access to private datasets to further AI development.
AI agents using blockchain wallets to support payments processes and commerce.

These examples reflect trends we are observing that explore how the convergence of AI and DLT could: make smart contracts better and more accessible; make blockchain processes 'smarter' in their use of real-world data; improve and enhance AI training and development with DLT-based controls for data holders offering immutable records of data provenance; and leverage digital identity, for example by taking the next steps towards frictionless, payment-enabled agentic AI.

Our hope is that, by contributing to the growing discussion on the potential of these technologies, this paper will help further define the issues for exploration and innovation.

Navigating the rapidly evolving legal landscape

The convergence of AI and DLT necessitates consideration of a complex patchwork of legal and regulatory requirements, including evolving technology-specific laws, data protection requirements, contract and private law considerations, as well as sector-specific regulations that could impact use cases for financial services, healthcare, energy, transportation and more. These parallel and overlapping regimes and laws demand careful, case-by-case legal analysis to ensure compliance and to mitigate liability and other risks. In some cases, possibilities for innovation may be in tension with frameworks for accountability and liability, particularly where any use cases move towards true decentralisation.

This paper spotlights some examples of key legal considerations relevant to AIxDLT. For example, the last few years have seen significant changes to the legal landscape for AI and DLT, including the introduction of AI-focused legislation. These laws have developed alongside new comprehensive regulatory regimes governing markets in cryptoassets and cryptoasset service providers. This trend towards more technology-focused regulation will continue apace as more jurisdictions adopt specific frameworks and regulatory approaches will likely become more sophisticated as use cases develop. In parallel, courts across the globe are starting to grapple with novel liability issues in relation to both AI and DLT technologies, which will inform how AIxDLT projects are structured and documented. In addition, AIxDLT use cases can give rise to issues around contractual interpretation and technical errors in the context of smart contracts – and associated liability risks and potential disputes – with considerations differing depending on whether a smart contract complements a conventional natural language contract with ''off-chain'' provisions or a code-only smart contract is used.

While there are numerous hurdles and complexities, for many organisations the potentially transformative opportunities arising from the combination of AI and DLT may significantly outweigh the challenges. The groundwork for this transformation is already being laid, and, for many use cases, their implementation will not be in the far-distant future but a shorter-term reality.

This paper has been co-written in collaboration with Deutsche Bank:

Clifford Chance: Diego Ballon Ossio, Rita Flakoll, Laura Nixon, Alex Balducci, Marc Benzler, Jane Chen, Steve Gatti, Jack Harris, Devika Kornbacher, Kimi Liu, Holger Lutz, Kate Scott, Herbert Swaniker, Charlotte Walker-Osborn

Deutsche Bank: Joy Adams, Sabih Behzad, Thomas Brophy, Tim Mason