AI gets mixed reviews – it’s either the key to curing cancer and addressing climate change or it is a technology that will wipe out jobs and possibly even human existence as we know it. Those in the latter camp might have felt vindicated recently when an AI safety researcher quit Anthropic, developer of the large language model Claude, with a cryptic warning that the “world is in peril”, but in most sectors, AI adoption is on the rise.  

There’s little doubt that AI is a tool that belongs strictly in its correct context or it will deliver misleading or harmful outputs (as can be seen in this study by the BBC and EBU). However, smart application of this emerging technology could determine the winners and losers of the future, not just based on output but also market confidence and investment. 

Our ‘De-mystifying Digitalisation and AI in Maritime’ campaign brought together  maritime business leaders and technology experts – including experts in data and AI working at digital leaders Google and Uber – to provide the crucial learnings from their work in data that would be applicable in a maritime setting. Here are some of the key considerations for our sector:

Data must be seen as a business asset 

Data hygiene underpins much of maritime’s ability to harness the full digitalisation potential. Our expert panel stressed the importance of a fundamental rethink of what data delivers – it shouldn’t be viewed simply as a byproduct of operations but as a source of genuine value. Companies that successfully combine the data they gather as part of their day-to-day operations and combine this with either commercially available or open source data stand to unlock significant opportunities, delivering value for their own organisation and across their industry. 

Successfully getting to this point means investing in Master Data Management – the less glamorous but vital foundation that involves harmonising data points, such as what constitutes a “vessel” or a “customer” across the organisation. It’s with this in place that businesses can establish a ‘single source of truth’ – with the entire organisation looking at the same agreed numbers to take data-informed decisions, as opposed to operating in silos using different systems and disparate data.  Furthermore, the data lake must be carefully maintained to prevent misinformation from creeping in which could amplify errors in outputs.

Change must be incentivised correctly

Overhauling internal culture to ensure data hygiene forms the foundation of effective AI use is unlikely to prove straightforward.  Roundtable participants unanimously agreed that maritime has historically been a conservative industry and that fully embracing the digitalisation potential could require an ‘asteroid moment’ – an existential threat such as that faced by physical retail with the growth of e-commerce – if it is to adapt to effectively embrace emerging technologies. 

This adaptation is likely to be driven by a combination of the ‘carrot’ – the commercial gains new technology offers businesses – and the ‘stick’, which encompasses regulations such as the Carbon Intensity Indicator (CII), which will force the industry to develop its understanding of performance data. 

Focus on the wider context 

Experts in the discussion highlighted various means by which AI was being harnessed to deliver outputs at breakneck speed, significantly boosting efficiency. These ranged from charter party agreements and new vessel designs that could be generated within a matter of minutes through to avatars of industry leaders to deliver messaging to teams (saving the leader’s time to focus on other work).

However, there was quite a fiery debate on how using pattern matching to generate these outputs magnified the impact of any errors introduced in contracts  – with one participant pointing out that maritime lawyers would see AI contracts not as a threat but as an opportunity to secure juicy cases with multiple defendants who would have been impacted. Additionally, speakers warned that using existing vessel designs to craft new ones was self defeating as AI was unlikely to come up with something truly novel – and also it did not have sufficient design information to safely design for vessels using new fuels.

Adopt a pragmatic approach
Almost as important as creating the culture that will allow a business to successfully embrace AI in advance of an ‘asteroid’, maritime businesses also need to understand when they shouldn’t use AI. This can be a challenge when the hype around AI is presenting it as a silver bullet to fell operational inefficiencies.  

One of our featured speakers expressed this point colourfully, noting that maritime businesses risk buying a Rolls-Royce when they only need a skateboard. To put it more plainly, there are instances where expensive AI solutions aren’t necessary and an automated spreadsheet is more than up to the task. The speaker was clear that having good data even in a basic system was far more valuable than an expensive system that had incoherent data.

Establish robust AI and data policy and governance

There was also a healthy debate around governance – which often drives organisational culture, particularly in a heavily regulated sector such as maritime, which is compliance focused. The lack of data standardisation in maritime is a well known issue, which has created roadblocks to the adoption of the maritime single window, complicated Just in time arrival and created data siloes that prevent interoperability.

Furthermore, the Intent Communications team has a large amount of anecdotal evidence that most organisations are exploring the use of AI on an ad-hoc basis without sufficient training in place, and a recent study has shown that very few maritime organisations have official AI policies and related governance. 

This lack of company policy or guidance introduces the risk that team members are inputting sensitive information into public AI tools without oversight. Our panel noted the importance of well-established policies that would prevent team members from copying unaudited financial figures or privileged legal information into public tools, such as ChatGPT. 

Future roundtables and project participation

Many of these points will be explored in greater detail at future roundtables and also in articles and video content on the Intent Communications website and social media. If you would like to nominate yourself to participate in the roundtables, please do reach out to info@intentcommunications.co.uk. Please note that seats are limited, so preference would be given to relevant contributors.

We also invite participation in the form of contributed comments and articles, video conversations and more. 

Sponsorship opportunities are available for organsiations that would like to be associated with this project. Please email namrata@intentcommunications.co.uk for more information and pricing.

We would like to thank our current sponsors for their support: Platinum Sponsor Bureau Veritas and Cobalt Sponsor Britannia P&I Club.