EU Parliament approves AI Act: Food manufacturers face smooth transition but consumers left exposed

14 Mar 2024 --- The European Parliament voted overwhelmingly in favor of the EU AI Act yesterday, with the groundbreaking regulation expected to enter into force early next year and apply in 2026. The impacts of this new law on food manufacturers will likely be relatively minimal as the industry’s current use of AI systems will mostly fall into the “Limited risk” category.

However, the European Consumer Organisation (BEUC) warns consumers remain insufficiently protected by the AI Act.

The new regulation, which is expected to shape future global attempts to manage the risks associated with AI systems, takes a risk-based approach across four categories: “Unacceptable risk,” “High risk,” “Limited risk” and “Minimal risk.”

“Companies need to prepare now and familiarize themselves with the AI Act to see if and how they will be affected — they need to determine their risk categories,” Julia Buech, a senior consumer foods analyst at Rabo Research Food & Agribusiness, tells Food Ingredients First.

“That said, food manufacturers may soon find that the implications for their business will be relatively limited as most AI systems that are currently in use will likely fall into the low-risk category.”

The “Limited risk” category mainly relates to the risks associated with a lack of transparency in AI usage. The regulation will introduce transparency obligations to ensure people are informed when using systems that carry the risk of manipulation, such as chatbots and image- or video-generating AI.

Consumer protection
The regulation, which was agreed in negotiations with member states in December 2023, was endorsed by MEPs with 523 votes in favor, 46 against and 49 abstentions. But BEUC says the result is underwhelming given the breadth of risks consumers will continue to face.

“Consumers rightly worry about the power and reach of AI and how it can lead to manipulation and discrimination, but the AI Act doesn’t sufficiently address these concerns. Too many issues have been left under-regulated, with an over-reliance on companies’ goodwill to self-regulate,” says Ursula Pachl, deputy director general at BEUC.

“There are, nevertheless, some important provisions that will allow consumers to take action if they have been treated unfairly or have been harmed. For example, they will be able to go to court as a group if an AI system has caused mass harm.”

Companies could face penalties ranging from €7.5 million (US$8.2 million) or 1.5% of global turnover to €35 million (US$38.3 million) or 7% of global turnover, depending on the type of violations.

Implementation issues
EU countries are set to approve the AI Act in May, but the law will not apply until 2026. This two-year implementation period creates problems, according to Nitish Mittal, partner at Everest Group, as the law could become outdated.

“Since the implementation might take nearly two years, it will need to be evolutionary to incorporate updates and innovation in AI during that time frame. How this will be tackled is not clear at this stage,” he tells Food Ingredients First.

“At the same time, there are concerns about how the AI Act might prevent exponential innovation in AI in Europe, compared to the strides being made by US and Chinese companies and governments.”EU companies are concerned the AI Act might slow innovation in AI, giving the US and China a competitive advantage.

The EU AI Act is the world’s first regulation to attempt to broadly manage the risks associated with this rapidly evolving technology. The law will influence governments in other regions as they prepare their own frameworks.

Meanwhile, BEUC wants to see other legislation work in combination with the AI Act to protect consumers.

“It is crucial that authorities make sure this legislation is properly enforced. Other legislation, such as product safety, GDPR and consumer law, must be used to provide an additional safety net for consumers. Consumers should never be left unprotected or powerless when AI is used to make decisions about them,” says Pachl.

Animal protection groups are also concerned that the incoming regulation fails to address the potential impacts of AI systems on animal welfare.

AI unleashed
AI holds the potential to transform the consumer food sector by refining operation processes, forging deeper connections with consumers and developing innovative solutions for sustainability and nutrition.

In a recent interview, Florian Schattenmann, Cargill’s CTO and VP of R&D and Innovation, highlighted some examples of how the food giant is harnessing AI, including a project to improve the health of chicken flocks.Consumers unprotected? BEUC claims too many risks have been left under-regulated by the new law.

“We have a project called Galleon that allows us to screen the DNA of the microbiome of a chicken flock — so a farmer can get samples of chicken poop from their farm, send it to us, and we run an analysis and tell them how healthy that flock is. The farmer can then adjust the feed to improve the health of the chicken flock,” he told us.

“Over time, as we get better data around different chicken farms, we can see there may be a lot of fluctuation in farm A and much more stability in farm B, so we can adjust and predict what the farmer will need. So, the model is learning by getting more data from more samples. A healthy flock will be more productive and have less mortality, and it will be better for the GHG footprint.”

Cargill is also bolstering its efforts to protect forests by leveraging satellite-powered risk monitoring capabilities across its soy, palm oil and cocoa supply chains.

However, the risks associated with AI systems are numerous. A recent Rabobank report identified affordability barriers driving a greater divide between big and small businesses as a major issue, as well as “AI washing.”

“AI is already very much part of life and business, so it is good to see that it will be part of legislation, too,” says Buech at Rabobank.

“The AI Act is not the end of the journey, but the starting point for a new model of governance that aims to be both solid and flexible as the technologies continue to rapidly develop.”

“While it seems that AI is faster than the law, the AI Act is certainly a landmark legal framework that has the world’s attention.”

By Joshua Poole