Ceramics industry and AI: opportunities and challenges in the face of the new European framework

Artificial intelligence ("AI") is not a matter reserved for large technology companies or innovation labs. In industrial sectors intensive in processes, design, logistics and quality control, such as the ceramic sector, AI is beginning to become an increasingly relevant tool to compete in demanding markets, reduce inefficiencies and anticipate operational problems.

In the ceramics industry, its potential is particularly evident. AI can be applied, for example, to quality control, furnace optimization, predictive maintenance, demand forecasting, inventory management, the development of new textures and designs, or even the adjustment of glazing processes and material formulation. These applications have already been identified by specialized entities in the sector itself, which point to AI as a technology with the capacity to impact manufacturing, distribution, design, maintenance and efficiency.

However, the same technology that improves productivity can also introduce new legal, technical and reputational risks if implemented without a minimum governance strategy. That is precisely one of the keys to the new European Regulation on Artificial Intelligence – Regulation (EU) 2024/1689 ("RIA") – which does not intend to stop innovation, but to order its use according to the level of risk that each AI system may generate.

From technological opportunity to legal obligation

The RIA is based on a simple idea: not all AI systems are the same and not all generate the same risks. For this reason, it establishes a tiered regulation model, based on the classification of AI at different levels of risk, which means that low-impact uses are not subject to disproportionate burdens, while those systems that may significantly affect the safety, health or fundamental rights of people must comply with stricter obligations.

In practice, a ceramic company can use AI in very different areas. It is not the same to use a predictive analysis system to optimize the energy consumption of a furnace as it is to use an AI tool to automatically evaluate the performance of workers, assign shifts, select candidates or monitor work behavior. Nor is it the same to use generative AI to support the creation of commercial proposals or catalogs as it is to incorporate a system that makes critical decisions about the conformity of a product or the traceability of a production line.

The first step, therefore, is not to prohibit or indiscriminately accept any AI tool, but to understand what is being used, for what purpose, with what data, by whom, and with what real impact on people, processes and business decisions.

Who is affected by the RIA?

The scope of application of the RIA is broad. The standard is not only aimed at those who develop AI systems. It can also affect importers, distributors and deployers, i.e. companies that use AI systems in the framework of their professional activity. Consequently, a ceramic company that incorporates an AI tool from an external provider is not exempt from regulation simply because it has not been developed internally.

This is particularly relevant in a sector where many technological solutions can be integrated into machinery, industrial software, planning platforms, marketing tools, customer management systems or human resources applications. AI can be present in a visible way, but also incorporated into solutions that the company already uses without initially identifying it as such.

The most sensitive uses for the sector

In the ceramic environment, many uses of AI are expected to be aimed at improving production processes, reducing defects, optimising energy, anticipating breakdowns or adjusting production to demand. These uses can be very beneficial and, in many cases, will not necessarily be considered high risk from the perspective of the RIA.

However, special attention should be paid to human resources. AI systems used in recruitment, promotion, performance evaluation, task assignment or worker monitoring may be classified as high-risk. The work environment is one of the areas in which the RIA places the greatest demands when AI can influence decisions about people.

It will also be necessary to carefully assess systems that may affect product safety. In these cases, the analysis should not be limited to the technology in the abstract, but to the concrete context of use.

Obligations: Beyond Buying a Technology Tool

One of the common mistakes in AI adoption is to consider that regulatory compliance falls exclusively on the technology provider. However, the RIA distributes responsibilities along the value chain. For user companies, there may be relevant obligations: use the system in accordance with the supplier's instructions, ensure adequate human supervision, keep records, monitor the operation of the system, report incidents and ensure an adequate level of AI literacy.

In addition, in certain cases, it will be necessary to comply with transparency obligations, particularly when people interact with AI systems or where synthetic content is generated.

A calendar that is already underway

Although the obligations are implemented progressively, the RIA calendar has already begun. From February 2025, prohibitions on unacceptable risk systems and the AI literacy obligation are enforceable. From August 2026, transparency obligations begin to apply. Finally, obligations relating to high-risk systems will apply from August 2027 and, in some cases, December 2028.

Therefore, adaptation to the RIA should not be approached as a last-minute reaction.

The risk of doing nothing

The RIA's sanctioning regime is one of its most relevant elements, as the most serious infringements can reach up to €35 million or 7% of annual global turnover.

But the risk is not only punitive. Poor AI implementation can lead to production errors, confidentiality issues, biased decisions, data exposure, supplier dependency and loss of trust from customers and distributors.

Roadmap for ceramic companies

The good news is that adaptation to the RIA can be approached in an orderly and proportional way. For a ceramic company, a reasonable roadmap could be structured in five steps.

First, identify all current and planned uses of AI. Second, classify each system according to its purpose and impact. Third, approve an internal AI policy defining permitted tools and prohibited uses. Fourth, establish internal governance, involving technology, legal, risk, operations, HR and business functions. Finally, implement periodic reviews, training and inventory updates.

Innovate with confidence

The ceramic sector has historically demonstrated an enormous capacity for technological adaptation, internationalisation and continuous improvement. AI can become a new lever for competitiveness.

But for this opportunity to be sustainable, it must be accompanied by control, traceability and accountability. The key question is no longer whether ceramic companies will use AI, but whether they will do so with sufficient governance to take advantage of its benefits without taking unnecessary risks.

The RIA should not be seen solely as a regulatory burden. Well managed, it can be become an opportunity to organize processes, strengthen customer and supplier confidence, reinforce legal certainty and differentiate companies in an increasingly technological market.