By Sara Lewis

Lately, I’ve been eager to get more informed about AI.

With several decades’ experience in science and traceability within the seafood world I have seen tech buzz words and “it” technologies come and go. But none quite as fast, or with as much potential to transform environmental and business contexts as AI. 

For decades, the global seafood sector was defined by its opacity.  At the start of my career in seafood, one of the biggest hurdles across the board was a fundamental lack of information. Ocean managers and conservationists were largely operating in the dark, with little visibility into human activity at sea or the complex journey of a fish from catch to consumer.

Today, the tide has turned. Globally, we have seen a staggering 30-fold increase in the volume of digital data over the last decade, powered by a surge in satellite coverage, the proliferation of Internet of Things (IoT) sensors in aquaculture, and the rise of smartphones. With the help of regulatory action, the emergence of standards and new monitoring tools, and efforts to digitize the historically paper-based processes within companies, we have moved parts of the industry from data deserts to a state of information abundance. 

At FishWise, data has always been a core part of what we do, and we have leaned way in on traceability, transparency, analytics, and machine learning. We are passionate about transforming information into sustainability action, via due diligence and adaptive management. All this work (and enthusiasm for collaboration) has helped us cultivate a fantastic network of data-minded experts.

So the moment feels ideal to start some conversations about AI with that community: where we’ve come from, what’s out there, and where AI may be heading for seafood companies. 

For this first of our three-part blog series on AI in the seafood sector, I spoke with Eric Enno Tamm, CEO and Co-founder of ThisFish Inc., and asked him to reflect on the evolution of data in seafood, and some of the enabling technologies that have paved the way for a more data-rich seafood industry.

Seafood has historically been a paper-based industry, relying on physical logs, clipboards, and disjointed Excel spreadsheets just like most traditional sectors. Transitioning this analog industry into the digital age required basic infrastructure first. As Eric describes it, tech innovators initially had to act as “digital plumbers”—laying the digital pipes to move data from paper to tablets on the factory floor. The last 15+ years of technological progress—driven by the ubiquity of cloud computing and mobile technology—has removed many economic and technical barriers. Eric identified the iPhone and AWS as major inflection points that made computing ubiquitous.

But not all tech developments have lived up to their hype – Eric noted that while there was significant buzz around Web 3.0 and blockchain technology in the late 2010s, most of its business value never materialized.

It is also critical to acknowledge that data and technology uptake is not happening at the same pace across the seafood industry.

There is a stark divergence in how readily data tools have been accepted across these domains. Currently, nearly 70% of aquaculture technology utilizes machine learning or AI, compared to only about 20% in wild-capture fisheries. Aquaculture continues to attract significant investment and adopts technology to improve the bottom line, whereas wild-capture fisheries remain largely data-poor and regulatory-driven.

This divergence comes down to an evolutionary split between incentives and regulations. Farmed operations have clear market motivations to adopt data tools because the insights directly optimize their two highest business costs: feeding and reducing stock mortalities.

In contrast, data tool adoption in wild-capture fisheries has historically been driven by regulatory compliance and electronic catch monitoring rather than immediate market dynamics. Until companies in wild capture see a clear ROI from implementing data technologies, they won’t, and we shouldn’t expect AI (or any tech solution) to resolve that on its own. And besides, as most of us within the seafood data world will acknowledge, the greatest remaining bottleneck is not technical, but behavioral, as the industry grapples with attitudinal shifts toward data sharing and transparency.

Small-scale operators also often face distinct hurdles, where foundational digitization remains difficult due to capital and infrastructure constraints, and we still can’t ‘magic’ our way out of some of the capacity barriers to uptake. For example, small-scale fishers can fail to adopt technology due to a range of reasons that are linked to values, culture, finances, power structures, and other issues that are beyond a tech fix.

Industrial operations, meanwhile, can create an overwhelming flood of raw data. ThisFish estimates that an average industrial tuna cannery processing 100 metric tonnes of raw material per day generates over 4 gigabytes of data a year—the equivalent of 2.7 million pages of text. Similarly in aquaculture, an industrial salmon operation may generate trillions of data points from feeding platforms, environmental sensors, and supply chain logs, but this information often lives in disconnected silos. Without a way to connect these fragments, transparency breaks down, and companies can be left with metrics without meaning.

So while areas of transparency remain a concern, the industry’s biggest challenge is no longer just getting the data. Instead, it is understanding the implications of that data and knowing how to respond. 

Meanwhile, AI tools are being marketed to seafood companies several ways, including:

• Translating Data into Actionable Insights
• Predicting Biological and Financial Risks
• Automating Complex Compliance and Monitoring Tasks

In the next blog, we’ll narrow in on the question: Is AI the missing intelligence layer in sustainable seafood?