Europe stands at the intersection for its future AI. Great plans are in motion to adopt the power of artificial intelligence, with the EU continent’s action plan and the upcoming Cloud and AA Development Act. However, in order for AI to work for all, the plan must be clearly delivered steps.
This blog looks at what is needed to build a truly A-edited continent Through five key steps, Europe must take AI, support innovations, guarantee safety and seize its workforce. Our knowledge is drawn from our leadership in AI infrastructure, network creation, security and cooperation and our obligation to international AI principles:
- Create the robust foundation AI: Safe, scalable, energy efficient.
- Secure AI: Cyber safety, risk management, global standards.
- Protect innovations: Safe development AI, global standards, evaluation of weaknesses AI.
- Keep data flow: Flexible data rules for AI training, reduced mandates of localization.
- Seize people with ai skills: Extensive training, partnership, modernized public procurement.
1. The robust buildings of the AI Foundation
AI thrives robust digital, physical and energy infrastructure. The desire of Europe to triple the capacity of its data centers in the next 5 to 7 years is a vital step along with the advancing AI and Gigafactories AI factories. This infrastructure must be safe, reliable and energy efficient.
Our research shows that only 7% of the European organization is considered to be prepared and quoted the limitation of infrastructure, security concerns and gaps of skills as the primary barrier. However, maximizing the AI throughout the economy requires trustworthy infrastructure and robust security.
Similarly, only 8% of EU companies feel that their infrastructure is ready for AI compared to 15% globally and 66% recognizes limited scalabibility. That’s a big gap. To add this, the EU must support strategic investments in new and modernized data centers, prefer security, durability and energy efficiency.
AI needs a connection. Wideband investment motifs, implementation of the Gigabit Infrastructure Infrastructure Act quickly and more efficient for regulatory regulatory construction of data centers and high -speed broadband deployment throughout Europe. The EU should also order the replacement or alleviation of outdated network assets to increase security and performance.
In addition, Europe urgently must strengthen its infrastructure through flexible funding mechanisms that support sovereign and global public cloud services. To gain access to the best technology in the world, a mixture of global and local solutions will give Europe the flexibility it needs. We have European success for his ability to access the best technologies.
Finlly must speed up energy acquisition and at the same time increase the efficiency of technological infrastructures supporting AI. This included support for the modernization of the grid, ensuring safety in the underlying control networks (including the solution of the fragmented NIS2 implementation) and the stimulation of the ICT device through schemes such as the Sustainability Scheme for data centers, with tax incentives.
2. Secure AI, from the bottom down
Recent development has shown that even the most demanding AI models can be deceived or abused. The attackers are increasingly focusing on the AI infrastructure and unique vulnerability in the AI deployment environment, from the compromises of the supply chain to the challenge of injection and data poisoning.
We must protect the technology that operates AI and the AI system itself, from how they were built to how they were used. When AI becomes the most obvious, it also expands the surface of the threat that can be used.
The EU should integrate strong cyber security policies into investment projects, which requires robust risk management. The key to strengthening AI security is adherence to internationally recognized standards such as Nist AI risk management, OWASP Top 10 and Miter Atlas.
AI can also be our strongest defense. Cyber safety-related cyber security can analyze huge amants, detect anomalies and automate the centers at the speed of the machine. The EU should support these cyber security tools and ensure that they are No Considered high risk according to the law of AI, thus supporting innovations in this critical area.
The use of generative AI (Genai) to increase cyber safety and expansion of cyber labor is another significant step as these tools democratize access to advanced security, simplify complex tasks and help solve the lack of cyber security experts.
Finally, support for regulatory alignment and mutual recognition of security certifications and conformity assessment will result in compliance with regulations and speed up the deployment of vital safety solutions and A.
3. Protect innovations to evoke new ideas
AI’s new ideas will not flourish without trust and security. Quick development and deployment of open source code, while accelerating AI, also pose potential risks, included data leak, data poisoning and uncertain outputs. An example of Deepseek R1, which showed 100% success in attempting to escape from prison in Cisco, emphasizes that even high -performance models can have significant security shortcomings.
It is essential to cooperate with the private AI model testing sector before and after deployment, encouraging an automated red team and continuous verification. Europe should support technological solutions that automatically test AI on the weaknesses, and to cooperate with international counterparts to promote receiving globally recognized standards and procedures in cyber security AI and risk management. Promotion of tools such as the taxonomy of opponent’s machine learning (AML) will help identify and alleviate the risks associated with AI.
The task of ENISA evaluate the weaknesses in publicly disturbed AI models, the publication of the findings and promotion of market solutions will be reinforced by the AI supply chain. This effort requires strong cooperation between government agencies and the private sector, as the rapid development of technology and sharing information will become important requirements.
4. Maintain data flow and standards clear
AI learns from data. To release the European potential of AI, especially in areas such as healthcare, data rules must be sufficiently flexible for AI training while protecting privacy. It will be crucial to ensure European data on AI revolution (eg AI training).
Data location often slows down innovation, costs more and can even create new security risks. With this in mind, the EU should strengthen its international data flow policy by monitoring multilateral data transfer agreements. These agreements should conclude mutual recognition of countries with robust data protection framework, unlike existing rules that have not provided long -term legal CEA, prevent business and innovation.
And as far as AI works, we should focus on internationally recognized standards. This means that everyone speaks the same technical language, which makes it easier for businesses to innovate and work smoothly for AI. This approach simplifies agreement for businesses and strengthens Europe’s attitude as a global leader.
5. Strengthen people with AI skills
AI fundamentally takes over industry and labor markets, so we have to make sure everyone has skills to prosper. Our research shows that only 9% of EU companies feel that they have the right talent for AI, which is a sharp contrast to 21% in the US.
We must invest in educational programs that teach AI skills. Partnership between governments and societies are vital. The EU must also encourage and anchor hiring skills in the private and public sector. Cisco is trying to train 1.5 million Europeans in digital skills by 2030 and train 5,000 instructors in AI and data science.
This is not just AI professionals – 92% of jobs Ict will be meaningful or slightly transformed by Genai. That is why Cisco, Twagging with eight leading global companies and advisors founded the consortium of the Ictr Dorkforce Ai-E-with-SE-Porad, analyzed how job profiles would affect AI, sharing knowledge and detailed training offers that help individuals reskill and upskill.
Support that digital transformation support in the public sector not only improvises services and security, but also promotes AI adoption and helps distribute data silos. It is important that the EU should modernize the public revenue process to associate with emerging technological solutions and refrain from the introduction of wide European preferential provisions, as they often limit the choice, prevent innovation and rarely provide experimented long -term local growth.
Can Europe become a continent of AI?
Europe has the opportunity to build the future of AI, which is smart, but also stop, fair and inclusive. It means to think about large, invest wisely and supports cooperation between the creators of politicians, industry and academy. By focusing on strong foundations, robust security, innovation, open data flows and qualified workforce, Europe can indeed become a continent of AI.
Cisco is a global leader in network solutions, security and cooperation. We have committed to supporting European AI Day. As a signatory of the EU Pact AI and the supporter of the international AI principles, we understand that maximizing the value of AI throughout the economy requires trustworthy infrastructure and robust security. Read our detallation recommendations for the AI continent Action Plan and the Cloud and AI Development Act here.
Share:
(Tagstotranslate) AI