• Menü menu
  • menu open menu
News 
Digital,General

Artificial Intelligence (AI)

Imagebild für Cybersecurity und AI

JOANNEUM RESEARCH is one of the central hubs of the Austrian AI community Credit: JOANNEUM RESEARCH/Bergmann

We have an exciting journey ahead of us, let's keep calm and take responsibility.
Branka Stojanovic

Are we SAVE_?

Early detection of Alzheimer's, new dimensions in robot autonomy and real-time language translation – artificial intelligence (AI) has potential in numerous spheres. It is already influencing research in countless different ways while opening up completely new possibilities that can be turned into reality at breathtaking speed. AI experts Branka Stojanovic and Andreas Windisch from DIGITAL share their insights.

How does AI improve efficiency in research and development?

Andreas Windisch: AI is already being used to increase efficiency in research and development processes. Taking software development as an example, tools can be used that significantly speed up coding. Smaller sub-problems, such as the ones we encounter in research and development work, can be processed more quickly and efficiently with the help of AI systems. Another example is the opportunity to use AI systems to interact with scientific studies directly. Once scientific studies have been “shown” to an AI system, the contents can be queried. In this particular area of application, there are already fully-trained AI systems in place which have been coached using over 200 million published studies so that they are now in a position to provide detailed information.

How does AI influence the speed of innovation and what are the challenges for a research organisation like JOANNEUM RESEARCH?

Windisch: As one of the leading high-tech companies in Austria, we’re very much involved with the latest technological developments – not just on the research side, but also directly, in terms of applications for state-of-the-art tools. These are also based on generative AI. Here, the challenges include the rapid pace of technological advancement, the complexity of the subject matter and the computing resources required. And that’s before we even get to data protection, and regulatory and ethical aspects. But thanks to our broad expertise in all of these areas, we are well placed to meet them.
We see ourselves as one of the central hubs of the Austrian AI community thanks to AI projects with customers and industry partners, international and national research projects, participation in standardisation committees in the field of AI, through training courses and lectures, as well as teaching at universities and FH universities of applied sciences. This topic is affecting society as a whole. It's very close to my heart and I would urge everyone who is interested in it to join us in the dialogue.

Branka Stojanovic: : AI is catalysing innovation in interdisciplinary research in digital technologies by automating tasks, speeding up decision-making and facilitating predictive analysis. While this increases efficiency, it also poses challenges, such as ensuring data privacy, addressing ethical considerations, overcoming skills gaps, managing integration complexities, and navigating regulatory compliance. Research organizations such as JOANNEUM RESEARCH have a challenging and important role to play in bridging some of these gaps, bringing basic research closer to end users while building and maintaining trust in the technology.

What role does AI play in ensuring competitiveness?

Stojanovic: AI is driving competitiveness in digital technology research and markets by accelerating innovation cycles, automating data analysis, and improving end-user satisfaction. Its ability to uncover patterns, predict trends and improve decision-making enables researchers to stay ahead in rapidly evolving fields, ensuring continued relevance, efficiency and breakthroughs in the dynamic digital landscape.

Who should engage with AI and why?

Windisch: This is an easy one to answer: everyone should engage with it. AI cuts across every section of society, and it is already having a major impact on our everyday lives in lots of different areas. Because of this, it’s essential that people familiarise themselves with the technology so that they are better placed to assess its potential implications, while at the same time ensuring that they know how to use it to their own advantage.

Will R & D still be possible without AI?

Stojanovic: The fact is that R&D is possible without AI, and historically significant advances have been made without AI technologies. However, integrating AI into R&D processes can increase efficiency, speed up data analysis and provide new insights. AI is a powerful tool, but its absence doesn't make R&D impossible; it simply means relying on other methods and technologies to achieve research goals. The choice depends on the specific requirements, objectives and available resources of the R&D project. However, given the convenience of AI, the real question is: will anyone opt for research without AI in the future?

Blessing or curse : is AI a double-edged sword in cyber security?

Stojanovic: AI in cybersecurity is definitely a double-edged sword. On the one hand, AI enables advanced threat detection, automates responses and improves overall security. But on the other, malicious actors can potentially exploit AI for sophisticated cyberattacks. The challenge is to stay ahead of the curve with responsible use of AI, robust defence strategies, and ethical considerations to ensure AI remains a blessing rather than a curse in the realm of cybersecurity. This emphasises again the importance of applied research and ensuring competitiveness through innovation and resilience to evolving challenges.

DIGITAL is part of the EUs ResilMesh project, which will use AI to counter cyberthreats. What expertise is DIGITAL contributing to the project?

Stojanovic: ResilMeshResilMesh, a three-year long Innovation Action project funded by the European Union, focuses on creating an advanced security toolset based on cyber situational awareness and to address the challenges of today's dispersed and diverse cybersecurity landscape. As a key partner in the ResilMesh initiative, DIGITAL is focused on using cutting-edge AI to detect advanced cyber threats in both IT and OT environments within industrial control systems. In addition, DIGITAL is actively developing a tool that uses state-of-the-art reinforcement learning techniques to automate security testing in these dynamic environments. DIGITAL is proud to be part of an initiative that is pushing the boundaries of cybersecurity innovation and research.

Can cyber attacks be predicted using state-of-the-art technology or does this require quantum computers?

Windisch: Quantum computers are a very exciting technology that we are looking at from various angles here at JOANNEUM RESEARCH. Right now, they are at a relatively early stage of development. But there is one possible future scenario under which quantum computers could end up making cryptographic protocols – i.e. encryption methods – vulnerable. A concerted effort is already under way to design new processes that will not be put at risk by advances in quantum computing. On the other side of the coin, some scenarios are already playing out in which quantum computing is able to underpin cybersecurity. Examples of this include the creation of tap-proof communication channels that permit the secure exchange of cryptographic keys between two parties.

Stojanovic: Although cutting-edge AI technology has made significant progress in predicting and mitigating cyber-attacks, it is important to note that the promising quantum computers are not essential for predicting cyber-attacks. Currently, cybersecurity prediction and prevention efforts rely on the foundations of conventional computing and advanced AI technologies, rather than quantum computing.

Should AI decision-making bodies be multidisciplinary?

Windisch: Since AI affects all areas of society, there are many other aspects besides the purely technical side of things that need to be taken into account when it comes to AI-related decision-making processes. These extend to all areas of our lives and all specialist disciplines: from legal and ethical considerations to social science viewpoints and economic and ecological matters. And it is for precisely for this reason that responsible citizens must engage with the topic and play a part in the discussion by sharing their own personal insights. Ultimately, we need to work together if we are to utilise the technology for the benefit of all.

Stojanovic: AI technology has moved on from being only a technical tool to having a huge impact on the everyday lives of a large number of people. This includes really diverse aspects, from technical support to medical advice and much more. Multidisciplinary AI decision-making bodies are essential for a holistic approach to addressing the diverse challenges of artificial intelligence. By bringing together expertise from different fields such as ethics, law, psychology and technology, these teams can ensure a more comprehensive understanding of the technical, ethical and societal implications of AI.

In 2023 AI research raised important ethical and societal questions, especially with regard to data protection, autonomy and social justice. Do you see any ethical dangers in the use of AI?

Windisch: AI definitely brings numerous dangers and carries potential for misuse. But especially in Europe, where we have made a public commitment to upholding human rights, our concerns centre on using the technology for the benefit of citizens. So this means that we also need to come up with solutions for European AI systems as far as dealing with ethical challenges or – more generally speaking – with potentially detrimental impacts on humans is concerned. This is also where the field of "digital humanism" enters the picture. It approaches technological developments and their impact on people from the joint perspectives of human rights, democratic values and diversity in society.

Is there anything that AI will never be able to do?

Windisch: Well, there are a lot of limitations that I could cite here, but I’m going to confine myself to the following: as they are right now – and as they are likely to develop in the foreseeable future – AI systems aren’t in a position to seek answers out of a thirst for knowledge. And it is this curiosity, which we can observe in every child and which we unfortunately lose far too quickly in adulthood, that makes us so human. It’s the guiding light behind many of mankind's greatest and most creative cultural and scientific achievements. Still, and I need to choose my words carefully here, because the really hard and fast limits are the ones imposed by the laws of nature. So we need to be cautious when we make categorical statements within these limits, particularly in the light of recent developments.

Stojanovic: Last year I would have said that AI would never be able to cook a home-made meal. Today, there are experiments with robotic arms cooking in AI-controlled kitchens. Today I have to say - we have an exciting journey ahead of us, let's keep calm and take responsibility, because AI cannot be held accountable.

 

Your contact

Mag. Dr. Andreas Windisch
Forschungsgruppenleitung
Newsletter
Register
Would you like to find out more about JOANNEUM RESEARCH
Keep up to date with all the latest research topics and make sure that you don’t miss out on any of our exciting events. You can unsubscribe at any time.

More news

Skip to content