DigitalIndustryResearch & Innovation

Foundation models, generative AI and IP

By Márton Nagy, Minister for Economy, Hungary

As the minister responsible for national economy I see artificial intelligence as a technological tool that can achieve unprecedented efficiency in almost any field in the economy and society.

I’m not alone with this insight.

Due to recent research findings the economic potential of AI is incredible, “conventional” and generative AI solutions could add the equivalent of 15.7 trillion USD to the global economy by the end of this decade.

AI in itself also generates a continuously growing market. Due to estimates[1] the total machine learning market was valued at $19.45 billion in 2022 and will reach $188.34 billion by 2030. The Natural language processing market will grow from $18.9 billion in 2023 to $68.1 billion in 2028. The global computer vision market was valued at $10.6 billion in 2019 and will reach $24.03 billion by 2027.

Generative AI has a decent and rapidly growing contribution to these numbers. In particular, Foundation Models, large-scale machine learning models trained on vast amounts of data, have become a cornerstone of modern AI development. These models underpin many of the generative AI systems that create content — text, images, code, and more — without direct human input. And while after Chat GPT was made available to the wide public there’s a certain hype surrounding this topic, this transformative technology holds immense potential for literally all sectors of the economy, ranging from healthcare to manufacturing.

It should also be mentioned that while one of the main concerns with regards to AI is that ‘it will take our jobs’ (which is – by the way – a rather false assumption since both surveys and experiences shows that AI also generates new workplaces with higher added value), a recent study[2] highlighted that close to 80% of the jobs in the U.S. economy could see at least 10% of their tasks done twice as quickly (with no loss in quality) via the use of generative AI. The survey also highlighted that these improvements were particularly noticeable among low-skilled/low-performer employees.

This clearly shows the potential of this technology – the ability to improve existing workflows and the performance of the workforce is an important asset in productivity and competitiveness.

Goldman Sachs estimates that over the next decade Generative AI in itself will be responsible for a 0.4 percentage point increase in GDP growth in the U.S., 0.3 points in other developed markets, and 0.2 points in emerging markets.

However, there’s a special relation between Foundation models and Generative AI and the creative industry. While the industry itself also benefits from the use of this technology – it can undoubtedly help the work of writers, visual designers and even music composers – it also imposes its own challenges and raises questions to be answered. In the traditional understanding of IP law, creators and inventors are human. But with AI systems capable of generating original works, the lines between human and machine-generated content blur. Who owns the output of an AI system? Does the creator of the algorithm, the operator, or the AI itself hold the rights? And who is liable when traditional IP rights are violated by generated content?

As the Minister for National Economy, I believe the creative industry is as integral to the economy and competitiveness of a nation as any other sector. The gross value added of the Hungarian creative industry doubled between 2010 and 2022, accounting for approximately 2-3 % of the gross value added for the national economy. This growth rate of the creative industry is the second highest in the line between sectors. The industry has the potential to drive overall economic growth as it develops products and services in conjunction with many other industries.

In addition, its cultural significance holds a special value that cannot be quantified in monetary terms or numbers and therefore requires special attention or protection.

My ministry has already consulted with industry stakeholders on this matter. They are not against the technology, quite the opposite; they are keen to adopt it. However, the rights of authors must be adequately safeguarded.

As the Minister for National Economy, it is not my job to resolve every legal aspect of this issue. However, I believe that through collective thinking and consultation with relevant parties, we must strive to find a solution to these issues as soon as possible.

This solution must ensure a peaceful coexistence with technology and allow for the reaping of its benefits for developers, innovators, and copyright holders alike.

Therefore, we have raised awareness on the importance of this topic both within the Hungarian government and at the relevant forums of the European Union (most notably at the newly formed AI Board responsible for the coordination of the implementation of the AI Act), and we encourage dialogue and joint thinking in this area.

Conclusion

Protecting European values and regaining European strategic sovereignty in the global technology market dominated be the US and China is probably the most important European challenge of this decade. Hungary believes that adaptation and uptake has a key importance in European competitiveness.

We believe that Europe has the potential to win this race, and the recent adoption of the AI Act with creating a level playing field for all actors of the market was an important step in this direction.

We also believe that with further consultations and joint efforts we can achieve a unified approach to the recent issues of AI and intellectual property rights ensuring both innovation and the protection of the rightsholders and the European cultural heritage, while – as it was the case with GDPR – also setting up a global trend for this matter.

 

[1] Századvég – Pre-research for the Hungarian AI Strategy

[2] Generative AI Impact – Survey by Google/Andrew McAfee (Andrew McAfee is Technology & Society’s inaugural fellow, a Principal Research Scientist at the MIT Sloan School of Management, and the co-founder and co-director of MIT’s Initiative on the Digital Economy.): https://storage.googleapis.com/gweb-uniblog-publish-prod/documents/Generally_Faster_-_The_Economic_Impact_of_Generative_AI.pdf