Private AI: balancing innovation with data privacy in Europe

By Fabrizio Garrone, Enterprise Solutions Director at Aruba S.p.A.

In recent years, Europe's digital landscape has experienced a surge in AI adoption and development. In February this year, the European Union announced a substantial boost of funding of over €176 million to invest in projects to accelerate the adoption of AI across continent, signalling a pivotal turn in the region's so-called 'Digital Decade'. Laid out by the European Commission (EC), the ‘Digital Decade’ plans aim to see Europe achieve a prosperous and human-centred digital future by 2030.

Promising businesses increased efficiency and productivity, this explosion of AI development in recent years can be put down to two main factors:

1. Data availability: A huge amount of digital data is now available in Europe, fuelling AI algorithms' learning and improvement.

2. Processing power: Increased innovation around processing power, such as machine learning, with widespread investment across Europe, has enabled this data to be processed quicker and more intelligently.

However, businesses, technology providers and regulatory bodies alike have grappled with concerns over resolving concerns around the need to access data to train algorithms and limits posed by sovereignty and privacy across Europe.

The data and privacy conundrum

With this growth in AI usage comes a balancing act faced by businesses adopting this technology. Companies need to consider where the data they are using to train AI models comes from and under what privacy laws it may fall. With Europe implementing some of the strictest data laws in the world, considerations around data sovereignty and privacy may limit the extent to which models can be trained.

If Europe's "Digital Decade" is, by nature, going to be "human-centred", then ethical considerations around data privacy will be crucial to meeting the EC's targets by 2030. This is where private AI comes in. 

Private AI: a secure solution for European businesses?

Combining the best of both worlds, private AI allows on-premises training and running of AI models on a company's own data. For European businesses, this has many benefits. Private AI systems can run off a business's confidential data in a private setting, avoiding the sharing of share reserved data outside the company. This enables companies to reap AI's productivity and efficiency benefits without running into problems regarding data privacy compliance and control. This is particularly useful for large companies and public institutions that have an abundance of data to train models and are also subject to extensive data privacy requirements. 

By prioritising data privacy and ethical AI development, businesses can gain a competitive advantage in the European market and help build customer trust. For example, private AI minimises the need for extensive data collection and storage, reducing the risk of breaches and the associated fines under GDPR, translating to lower compliance costs. By keeping data use transparent and localised, European businesses can also demonstrate adherence to privacy laws more efficiently, avoiding complex procedures for cross-border data transfer. 

Europe as a leader in private AI adoption

With its focus on data protection, Europe is likely to be at the forefront of private AI adoption. Since GDPR imposes strict regulations on data collection, storage, and usage, private AI, by keeping data under the control of the business, aligns well with regulations. This makes it a natural choice for European companies seeking to comply. 

A recent report predicts the significant growth of the European data centre market in 2024, with a 16%  increase in supply in the core FLAP-D markets and 49% increase in market size of secondary markets. With this growth in 'home-grown" data centres, Europe stands in a solid position to host an influx of private AI infrastructures with new data centres not only located in FLAP-D region (Frankfurt, London, Amsterdam, Paris and Dublin) but also, and mainly, outside it. For instance in countries like Italy, European operators are driving growth as the data centre market is expected to increase from 411.4 MW in 2024 to 805.2 MW by 2029. With this projected increase in data centres across Europe more companies will be able to host their systems, including private AI, within Europe, aiding regulatory compliance and ensuring local needs are paid attention to. 

The future of private AI

With Europe's "Digital Decade" well underway, AI adoption will only increase. However, as the European Commission continues to focus on data sovereignty, ethical AI use will be at the forefront of regulation. Announced earlier this year, for example, the AI Act assesses the risks AI may pose and aims to set out measures to develop ‘trustworthy’ AI. Businesses looking to harness the power of this emerging technology in Europe, therefore, need to do so in a manner that will protect and appeal to European consumers. Innovations in private AI are likely to grow, offering businesses a competitive advantage and driving European customer loyalty. 

Frank Horenberg, Zivver’s Head of IT, and Simon Newman, Co-Founder of Cyber London, do a deep dive on the issue of investment in cybersecurity...
By Nicholas Borsotto, WW AI Business Lead and Head of Lenovo AI Innovators Program.
By Louise Lahiff, Chief Operating Officer, Version 1.

Preparing the IT Channel for an AI-Driven Future

Posted 1 day ago by Phil Alsop
BY Francis O'Haire, Group CTO, Climb.
By Suraj Tiwari, Global Head – Information Security, VFS Global.
By Marc Potter, CEO, Actian.
By Lars Rensing, CEO of Protokol.
By Aleksi Helakari, Head of Technical Office, EMEA - Spirent.