Some might say the past few months have been challenging for advocates of Artificial Intelligence (AI). Incidents ranging from crashes involving autonomous vehicles to dubious decisions by automated algorithms appear to have tipped the AI hype cycle into the “trough of disillusionment” phase. And more generally, people remain sceptical of “black box” AI whose reasoning lacks transparency and explainability.
Given all this, it could be easy to assume there’s a crisis of public confidence and trust in AI – not least around its use in public services. But you’d be wrong. Accenture’s recent Citizen Survey across six countries found that 50% of citizens support the use of AI in public service, and that support rises noticeably when presented with concrete benefits.
Border agencies should bear these findings in mind when planning out and executing their AI strategies. In my view, the message they send is clear: don’t be daunted or overwhelmed by the latest swings in the AI frontier – whether upwards or onwards – or let these mask the valuable opportunities that this transformational technology offers at the border.
What kind of opportunities? As cross-border trade and travel value chains become increasingly digitised, I think five value drivers will maximise the benefits of AI for border agencies. Here they are.
- Responsible AI. People must become confident that AI decisions are ethical and trustworthy. The good news is that agencies are already taking this imperative on board. According to the border perspective drawn from the Accenture Technology Vision 2018 , 78 per cent of public service executives say they’re seeking to gain citizens’ trust by being transparent in their AI decisions. Also, countries such as the UK and UAE have set up policy groups looking proactively at what AI can deliver.
- Intelligent automation. As I’ve already mentioned, most automation currently used at the border handles mundane tasks rather than those requiring cognitive intelligence. However, advances in AI will result in automation becoming increasingly intelligent and see it applied to an increasing array of applications. These might range from creating a chatbot to answer complex questions about duties on specific goods to developing entirely autonomous ports where humans focus on a monitoring role.
- Enhanced judgements. It will be many years before AI is trusted on its own to make decisions on issues such as whether to allow certain goods or visitors into a country. But AI does have a valuable role to play in augmenting human judgement and supporting choices about the “next best action” to take on casework. For example, it can make informed suggestions that enhance a caseworker’s decision-making, by pulling together and learning from all the data about a particular case and the compliance history of similar cases. This can result in better insights around the likelihood – for example – that drugs are being carried, leading to more accurate and efficient decisions around enforcement activities. However, it’s important to stress that the human caseworker remains ultimately in control.
- Enhanced interactions. One of the biggest benefits of digitisation is the ability to stop using paper forms and provide more personalised online and in-person service. With AI, personalisation can be elevated to a whole new level – and in an environment like customs, which is all about form-filling, the impact on user experiences can be transformational. A border agency’s wealth of existing and historical data about each transaction or entity means routine information-gathering can be almost completely automated, enabling the agency to focus more effort and energy on providing a positive experience and facilitating safe and frictionless passage of people and goods.
- Intelligent product categorisation. Customs classifications processes are notoriously complex and bureaucratic, and the descriptions used bear little relation to everyday language. Yet any entity who gets the classification wrong can face severe penalties. AI has huge potential for simplifying these complex nomenclatures, making it easier to find the right classification codes to use and simply understood by all parties, reducing costly errors, and fostering the effective flow of trade.
It’s in the nature of emerging technology to have ups and downs. But whatever the latest headlines may say, the genuine promise and business case for AI at the border remain as compelling as ever. So it’s vital that border agencies stay focused on the value that AI can deliver – and don’t let short-term concerns distract them from realising it in the years to come.