No matter what you think of artificial intelligence (AI) – revolutionary technology or overblown fad – it seems clear that, for better or worse, it is set to fundamentally shape the next decades of finance. Even more so in the trade sector, where the drive for digitisation –  together with excitement around AI – is making tradetech one of the most vibrant and growing fintech industries.

At the 2025 Sibos conference in Frankfurt, Germany, Trade Finance Global (TFG) sat down with a range of tradetech leaders at IBM, Cleareye.ai, Trademo, and Premium Technology to find out more about the inner workings of tradetech innovation and where it is headed. 

Beyond paper?

Despite vast advancements in other areas of global trade, documentation still lags woefully behind: only 7.5% of all global trade players use solely electronic bills of lading (eBLs), and over half are still exclusively paper-based documents. While eBL adoption is increasing worldwide compared to previous years, in 2024, 76% of paper-only users had no plans to switch to eBLs in the next two years. 

“[Trade digitisation] is something people have been talking about for probably 20 years by now, using general ledger or blockchain – but has just not been adopted,” confirmed Christoffer Eriksson, Global Partnership and Sales Head at Cleareye.ai. 

A large part of the issue is standardisation, explained Eriksson. Despite the ICC Digital Standards Initiative’s (DSI) framework for supply chain digitisation, published last year, corporates are still hesitant to adopt a technology that might not be recognised equally in all regions and may lack compliance with existing processes. This is making the industry painfully slow: processing a paper bill of lading takes 16.4 hours on average, according to a McKinsey report. Over thousands of shipments and dozens of trips a year, this amounts to an enormous time expenditure. 

Besides the processing times, paper documents also pose a significant risk for trade finance lenders. A range of trade finance frauds is conducted by providing banks with falsified bills of lading, or providing authentic ones to multiple banks to gain access to more financing. 

With eBLs, on the other hand, verifying documents becomes much easier. “Just like we have a vehicle identification number (VIN) number on a car, so that if my car gets stolen, I can look up the VIN number on the vehicle to see if this is still my car, even if it got painted a different colour. It’s the same idea. If you have these identifiers, you can start tracing documents in a much better way,” said Robert Saba, Associate Partner, Financial Services, Strategy & Transformation at IBM. 

How much can AI help?

While technology companies can’t shoehorn trade digitisation, they can offer other ways to optimise – for example, through AI. “EBLs might be the future, but it might take 20 more years until they get universally adopted or just until adoption increases. Until then, AI can generate a lot more value immediately in every trade process,” said Shalab Singhal, CEO and founder of Trademo. 

This doesn’t mean AI is replacing humans one-to-one – a prospect likely to horrify regulators and politicians alike, at least in the near future – but that it can be used as a booster to increase productivity. “It’s not like AI is replacing existing processes; it’s just helping define them, and means that tasks that were previously very manual can now be automated,” said Eriksson. 

For processes like KYC, for example, this could look like an AI extracting information on a company from the internet, and having human employees just verify that information instead of researching it themselves, said Michelle Leung, Product Manager at Premium Technology.  On the customer side, AI chatbots can help businesses and suppliers navigate new banking systems, which is especially crucial for smaller businesses that may not be familiar with trade finance’s hyper-specialised and often confusing terminology. 

Superpowers: How much value lies in data?

The real potential in AI, however, lies in its data processing capabilities. An AI model can take in extraordinary amounts of data on a bank’s customers and transactions – such as the type and cadence of financing requests, the terms agreed, their due dates and available credit – and give banks unprecedented insights into future movements. 

The same is true on a macro scale: AI models can examine complex trade relationships and come up with a multi-tier supply chain mapping that helps corporates understand the risks in their supply chains. Before AI, the same level of insight might have been restricted to one specific supply chain and region, and would have taken an extremely skilled analyst. 

“We have built a global trade knowledge graph, which is trained on shipping data, global trade, and transactional data coming in from customs parties across the world,” including tariff and export controls, sanctions, and corporate ownership registries, said Singhal. 

For example, the AI in IBM’s Connected Trade Platform can process all that information and alert companies when their supply chains might be affected by a shock or be exposed to compliance risks. 

—

With tech implementation, as with almost all aspects of global trade, everything is connected. “You have to address every layer of the stack, from the infrastructure, the cloud, the software layer, the management layer, to the governance layer. Otherwise, you’re just solving a tiny bit of the problem. It doesn’t give you the value end-to-end because often these things are all interrelated,” said Prakash Pattni, Managing Director of Financial Services Transformation at IBM. 

Be it through AI, trade document digitisation, or intelligent data platforms, the future of trade innovation lies in interoperability. Tradetech in the next quarter-century will see the industry finally evolve past paper and harness the full potential of AI, bolstered by technology providers who work together to enable frictionless innovation.