The Austrian State Secretary for Digitalization and Telecommunication Florian Tursky gave iGlobenews an exclusive interview. Tursky who is responsible for AI in the Austrian government, discussed the new Digital Austria Act, how the EU is dealing with AI regulation, how to protect data privacy while regulating and using AI and what the EU and the Austrian government is doing to protect against the dangers associated with AI. Tursky believes that Austria and the EU is taking all necessary precautions to ensure that AI is used safely, while supporting innovations.
Diana Mautner Markhof, 11 July 2023
iGlobenews: What is your opinion on the EU Parliament draft rules for AI and how will these rules apply to Austria regulating AI. Do these draft AI rules include sufficient protections and transparency and can they prevent AI influencing democratic processes and AI use for illicit/dangerous purposes.
State Secretary Tursky: Before I go into this question, I would like to briefly explain the procedure. It is important to mention that a general approach was adopted at Council level at the end of last year. This general approach represents a compromise agreed by the governments of all Member States. Despite some concerns about data protection and consumer protection, Austria agreed to the text by submitting a protocol declaration. The aim was to reach an overall compromise, which will now be introduced into the trilogue negotiations. The European Parliament has now adopted its position in mid-June. And will defend this position in the upcoming trilogue negotiations.
At national level, we are currently coordinating an interdepartmental position on the demands of the European Parliament. In addition, at the national level there is the joint working group “AI Policy Forum”, which was established in November 2021 under the joint leadership of the Ministry of Finance and the Ministry of Climate Action, Environment, Energy, Mobility, Innovation and Technology to accompany the interdepartmental implementation of the national AI strategy.
iGlobenews: Will the EU AI Office monitor AI implementation in the EU and how?
State Secretary Tursky: The position of the Council provides for the establishment of an EU AI Intelligence Board whose tasks are similar to the ones now proposed for the EU AI Office by the European Parliament. The foreseen tasks of the EU AI Office that shall be an independent body consist in supporting the implementation and harmonized application of the AI Act, providing guidance and coordinating joint cross-border investigations. It is important to stress that all tasks foreseen for the EU AI office regarding the application and implementation of the regulation are purely of a supportive and advisory nature and are without prejudice to the tasks of national supervisory authorities.
iGlobenews: The Austrian government passed the Digital Austria Act on June 1. What are the main goals of this Act and how will the government ensure that private information is not hacked and misused?
State Secretary Tursky: The Digital Austria Act (DAA) is an important milestone for Austria’s digitization and supports the federal government’s goals of maintaining and expanding prosperity and security through digitization. With 117 measures and 36 digitization principles, the DAA provides a solid foundation for future-oriented digitization that affects all areas of people’s lives.
Across departments, the measures take applicable data protection principles and barrier-free accessibility into account. For example, the Digi-Check is an important aspect of the review of laws to check whether they are suitable for digitization. This approach offers the possibility of making public services more effective and providing better support for citizens. We also ensure, by the Data Protection Impact Assessment (DPIA) according to Art. 35 GDPR, that all necessary security precautions are taken and that the personal data of the users are protected in the best possible way.
We also have a strong interest in using artificial intelligence (AI) responsibly and strengthening trust in this technology. As part of this effort, we are working on effective regulation in Europe through the AI Act, as well as at the national level through a future AI authority. These measures should help drive innovation and take the AI industry to a new level.
iGlobenews: Elon Musk has stated re OpenAI, the company that developed ChatGPT: “Open AI is training the AI to lie”, “to be politically correct” and “potentially manipulate public opinions”. Do you agree with this, and if not, why?
State Secretary Tursky: The advances in technology are remarkable and offer many opportunities for positive developments. However, we should always keep the risks in mind and work for transparency and data protection to safeguard fundamental rights.
At the same time, we must not forget that innovation is crucial for our global competitiveness. We should therefore adopt a balanced approach that takes into account both the protection of fundamental rights and the promotion of innovation. I am convinced that the new regulations for artificial intelligence will set a milestone, while taking into consideration that one must always take into account that the EU consists of independent states and it is necessary to find an overall compromise.
iGlobenews: A Federal judge in Texas has limited the use of AI in writing court filings and insisted that a human factcheck what AI has written. Is this a model on how to use AI and guarantee that AI bots are not writing court documents and student essays/examinations? If not, will this inevitably lead to a culture which does not see the need to learn, think and write, and is thus more easily manipulated.
State Secretary Tursky: In the Artificial Intelligence Act it is foreseen that providers of AI systems which fall under the category of high risk have to fulfil certain requirements and obligations before being allowed to place that AI system on the market. One of those requirements is that high-risk AI systems have to be designed and developed in such a way that they can be effectively overseen by natural persons during the period in which the system is in use. Taking the example from Texas, for the use of such a system within the EU it would be a requirement to have effective human oversight during the time of its use.