Accountability Doesn’t Change

This interview is part of our extended Q&A series exploring AI and food safety.

As AI becomes more embedded in food safety systems, questions about responsibility come into sharper focus. While these tools can support better decision-making, they do not change who is accountable when a product causes harm. In this Q&A, Bill Marler, managing partner of Marler Clark, explains how AI fits into existing legal frameworks and why human oversight remains essential.

Could AI tools increase legal exposure when it comes to food safety systems?

Yes—but it depends on where the entity sits in the chain of distribution. Liability in food cases always depends on that position. For example, if someone is a manufacturer—a restaurant making a hamburger or a processor handling lettuce—they are considered manufacturers under the law and are strictly liable.

What that means is I only need to prove that the product was defective—say, it contained a pathogen that caused illness or death. I don’t need to prove whether they were careful or careless, whether they used AI or didn’t use AI. None of that matters. I just have to prove that the product caused harm.

That’s different from other entities in the chain, like retailers or distributors. If a retailer simply sells a sealed can of peas, you have to prove they breached some duty [for example], maybe the can was dented or past its shelf life. But if that same retailer is preparing fresh-cut produce, then they’re functioning more like a manufacturer and are treated differently.

So, when you think about AI, whether a company uses it well, poorly, or not at all and if you’re a manufacturer and something goes wrong, it doesn’t really matter. Liability still attaches. Where it may matter more is with retailers or pass-through entities—companies that don’t manufacture the product—because then questions about oversight and duty of care come into play.

I don’t need to prove whether you used AI—I just have to prove the product caused harm.

Bill Marler

Marler Clark

How do you personally think about AI in the context of food safety?

I tend to think of AI as an additional tool to help make food safer. But like anything, my concern is that people will get lazy and rely on AI for answers without applying their own judgment.

As AI improves, it’s probably going to give better answers than humans a large percentage of the time, maybe 95% of the time. But the problem is that remaining 5%. We’re already seeing misuse in the legal field, where people are submitting briefs generated by AI that cite cases that don’t exist. Courts are catching it, and it’s a real issue.

You can easily imagine that translating into food safety, using AI to review inspections, analyze data, or interpret imaging like X-rays. These systems might be right most of the time, but if people rely on them blindly and remove that human layer of review, that’s where problems can arise.

Where do you see AI adding the most value in food safety?

If AI is used as a tool, it can do a couple of important things. First, it can help prevent outbreaks. And if you prevent outbreaks, manufacturers won’t get sued because there won’t be an outbreak in the first place.

Second, I think AI will become part of the standard of care. Retailers and others in the chain of distribution will likely be expected to use these kinds of tools going forward. If they’re not using them, it could eventually be seen as falling below the standard of care.

Looking ahead, what role do you expect AI to play in food safety programs over the next five years?

I think AI is going to become an integral part of everything we do. That’s really just human nature—we push the envelope on technology, even if we don’t fully understand the long-term consequences. The bigger question is whether what we’re building will ultimately be beneficial for us.

In more practical terms, AI is going to be deeply tied to things like big data and traceability. Analyzing massive datasets allows you to identify patterns and probabilities in ways humans simply can’t. If you flip a coin 10,000 times, you expect a 50/50 split. But if you start seeing 9,999 outcomes go one way, that tells you something meaningful—and that’s the kind of insight AI can provide.

In food safety, that means better traceability, faster identification of risks, and more informed decision-making. AI could make technologies like blockchain look primitive by comparison. The possibilities are pretty much endless in terms of how it could benefit food safety.

What risks or concerns do you have about AI becoming more integrated into food safety systems?

The biggest risk is losing the human checkpoint. There still needs to be some level of human oversight, someone who can step back and think critically about what the system is telling them. Otherwise, we risk over-relying on automation. And the reality is, humans can be pretty lazy. If a system appears to be working, there’s a tendency to trust it without questioning it. That’s where problems can arise.

How quickly do you think AI capabilities are evolving, and what does that mean for the future?

The pace of change is remarkable. I was talking to a friend who has ALS, and he used an AI system to record his voice before losing the ability to speak. Now he can communicate using a synthesized version of his own voice. That’s an incredible application of the technology. So, there’s clearly a lot of good that can come from AI.

At the same time, you see developments like advanced robotics that are evolving incredibly fast. It raises bigger questions about where all of this is heading. We’re building systems that are becoming more capable very quickly, and while that’s exciting, it also introduces uncertainty about how those systems will ultimately be used—and how much control we’ll retain over them.

What is the key takeaway for food companies adopting AI in food safety systems?

AI should be viewed as a powerful tool but not a replacement for human judgment. From a legal standpoint, responsibility doesn’t change. If a product causes harm, liability follows, regardless of whether AI was involved. From an operational standpoint, AI can help prevent problems and improve systems—but only if it’s used thoughtfully, with appropriate oversight.

The opportunity is real. But so is the risk if companies rely on it without maintaining that human layer of accountability.

This interview has been edited for clarity and brevity.