Among the many AI-focused sessions, the discussions around longevity, personalised medicine and data diversity were particularly relevant. These topics sit at the intersection of Healthcare AI, Explainable AI (XAI) and responsible AI development, raising a critical question for both researchers and technology providers: can AI systems be truly effective if key populations are missing from the data used to train them?

The seventh edition of Barcelona New Economy Week (BNEW) took place on 5-7 October 2026 at DFactory Barcelona. Organised by the Consorci de la Zona Franca de Barcelona, the event brought together hundreds of experts, innovators and startups across eight sectors: Artificial Intelligence, Health, Energy, Agrofood, Aviation & Space, Science, Talent and Gaming.

This year, Artificial Intelligence had its own dedicated track for the first time, after six editions in which it featured as a cross-cutting theme. The sessions explored how AI is reshaping employment, business models and decision-making processes, with contributions from organisations including the Barcelona Supercomputing Center, Oracle and Ibercaja.

The influence of AI extended far beyond the dedicated AI track. The Health programme showcased automation and live surgical robotics. Energy sessions focused on the growing role of data and AI in the energy transition. Across the venue, humanoid robots, collaborative robotics and autonomous mobility simulators demonstrated how quickly AI is moving from research environments into real-world applications.

For Mosaic Factor, BNEW provided a valuable overview of where AI innovation is heading across multiple industries. Many of the topics discussed, including trustworthy AI, data quality, robotics, predictive analytics and digital twins, directly connect with the challenges we address in our own projects.

AI Trends Observed at BNEW 2026

Across AI, Health, Energy and Industry sessions, several common themes emerged:

• Increasing adoption of AI-driven decision support systems.
• Growing demand for explainable and trustworthy AI.
• Greater use of predictive analytics and digital twins.
• Integration of multimodal data to support personalised services.
• Strong focus on governance, transparency and AI regulation.

These trends closely align with the solutions Mosaic Factor develops across healthcare, mobility and industrial environments.

While AI innovation was visible throughout the event, healthcare stood out as one of the areas where AI is expected to generate the greatest impact, while also raising some of the most important questions around reliability, explainability and bias.

Healthcare AI, longevity research and precision medicine

On 6th October, the BNEW Health session “Longevity and Healthy Aging: Strategies for a Longer-Lived Society” brought together researchers and clinicians working in genetics, neuroscience and integrative nutrition.

Beyond longevity itself, the discussion highlighted broader trends that are shaping the future of Healthcare AI:

• The transition towards precision medicine.
• The integration of multimodal data sources.
• The importance of data diversity in both clinical research and machine learning systems.

The panel addressed three developments with direct implications for Healthcare AI:

1. Epigenetics: lifestyle and environment influence how genes are expressed.
2. Personalised, multimodal treatment: combining interventions tailored to each individual.
3. Data representativeness: the need for more diverse research populations and datasets.

Each of these challenges is closely linked to Mosaic Factor’s work in predictive analytics and clinical decision-support systems, where model reliability depends on the quality, diversity, and representativeness of the data available.

Genetics and Epigenetics: how lifestyle shapes gene expression

One of the panellists used a card game analogy to explain the relationship between genetics and epigenetics. Every individual is dealt a hand at birth, inheriting half of their genes from each parent. Those cards cannot be changed. How they are played can.

Epigenetics describes how environmental and lifestyle factors influence whether certain genes are activated or remain silent. Physical activity, nutrition, sleep quality and stress management all play an important role in shaping long-term health outcomes.

The discussion noted that a substantial proportion of dementia cases is associated with modifiable risk factors. Research increasingly shows that healthy ageing cannot be explained by genetics alone. Instead, it emerges from the interaction between genetic predispositions, lifestyle choices and environmental conditions throughout life.

For Healthcare AI, this reinforces the need for models capable of integrating multiple sources of information rather than relying on a single variable or dataset.

Personalised, multimodal treatment in healthy ageing

The panel stressed that no single intervention can address ageing effectively. The most promising strategies are multimodal, combining different therapeutic approaches and adapting them to the characteristics of each individual.

Biological sex, ancestry, age and clinical history all influence how patients respond to diseases and treatments. As a result, longevity research increasingly overlaps with the principles of precision medicine, which aims to deliver the right intervention to the right patient at the right time.

This same paradigm shift is transforming Healthcare AI. Modern AI systems must increasingly integrate clinical, genomic, imaging and behavioural data to support more personalised and meaningful recommendations.

Healthcare AI Bias: why data diversity matters

The most significant takeaway for Mosaic Factor concerned the composition of the datasets that underpin both clinical research and AI models.

According to the speakers, much of today’s knowledge about ageing is still derived from studies involving people of European ancestry. Other populations remain comparatively under-represented despite significant genetic diversity. This has practical consequences: treatments and interventions that work well in one population may not have the same outcomes in another.

The same challenge applies directly to Healthcare AI.

A machine learning model can only learn from the data on which it is trained. If certain populations are under-represented, a model may achieve strong average performance while being systematically less accurate for specific groups. In healthcare settings, this can contribute to unequal clinical outcomes and reduced trust in AI systems.

As AI becomes increasingly integrated into clinical workflows, ensuring data quality and representativeness is therefore not only a technical challenge but also an ethical and regulatory requirement.

How Mosaic Factor builds trustworthy healthcare AI

At Mosaic Factor, trustworthy Healthcare AI begins with data quality, explainability and rigorous validation. As healthcare organisations increasingly adopt AI-powered decision-support systems, reducing bias and ensuring transparency become essential for both clinical trust and regulatory compliance.

Our approach is built around five core practices:

1. Assessing representativeness from the start. Before developing a model, we analyse who is included in the dataset and who is missing, considering factors such as sex, age, ancestry and clinical profile.
2. Validating performance across subgroups. Overall accuracy is not enough. We evaluate how models perform for different populations to identify potential disparities before deployment.
3. Explaining every prediction. Through Explainable AI (XAI), healthcare professionals can understand which factors contributed most to a prediction and incorporate that insight into their clinical judgement.
4. Communicating uncertainty. We seek to quantify and communicate confidence levels in model outputs. This principle also guides our work beyond healthcare, including Energy & Range Forecasting applications developed within the Twin-Loop project. AI systems should be able to indicate when they are operating outside the conditions represented in their training data.
5. Collaborating with clinical and research partners. Working closely with hospitals, research centres and clinicians helps ensure that AI solutions remain relevant, validated and centred on human expertise.

Precision medicine and responsible AI: a shared challenge

BNEW 2026 reinforced a message that was visible across the event’s AI, Health and Industry tracks: organisations are moving beyond exploring what AI can do towards deploying AI systems that must be reliable, explainable and capable of operating in complex real-world environments.

For healthcare, this means combining the promise of precision medicine with rigorous approaches to data quality, transparency and bias mitigation. The same foundations that support better research also support better AI.

The future of Healthcare AI will depend not only on more powerful algorithms, but also on better data, greater transparency and stronger collaboration between clinicians, researchers and technology providers.

At Mosaic Factor, these principles guide the design and deployment of our Healthcare AI solutions, from predictive analytics and decision-support tools to Explainable AI applications developed together with healthcare and research organisations.

Events such as BNEW help us stay connected to the latest AI innovations, emerging regulations and scientific advances that will shape the future of data-driven healthcare.

We thank the Consorci de la Zona Franca de Barcelona and all the speakers for an insightful event, and we look forward to continuing these conversations as AI adoption accelerates across healthcare and other critical industries.