Denmark on Wednesday laid out a framework that might help EU member states use generative artificial intelligence in accordance with the European Union's strict latest AI law – and Microsoftis already on board.
A government-backed alliance of major Danish firms led by IT consultancy Netcompany has released the white paper “Responsible use of AI assistants in the public and private sectors”, a blueprint that sets out “best practice examples” to be used Companies contain and support employees in using AI systems in a regulated environment.
The guide also goals to advertise the supply of “safe and reliable services” by businesses to consumers. The Danish Digital Government Agency, the country's Central Business Register (CVR) and pension authority ATP are among the many founding partners which have adopted the framework.
This includes policies that govern how the private and non-private sectors work together, deploy AI in society, comply with each the AI Act and the General Data Protection Regulation (GDPR), mitigate risk and reduce bias, scale AI implementation, data store safely and train staff.
André Rogaczewski, CEO of Netcompany, said the provisions set out within the white paper are aimed primarily at firms in highly regulated industries, similar to financial services. He told CNBC that he wanted to deal with a key query: “How can we scale the responsible use of AI?”
What is the EU AI law?
The EU AI Law is a groundbreaking piece of laws designed to control the best way firms develop, use and apply AI. It got here into force in August after previously receiving final approval from EU member states, lawmakers and the European Commission – the EU's executive body – in May.
The law applies a risk-based approach to regulating AI, meaning that different applications of the technology are treated in another way depending on the extent of risk. It is touted because the world's first major AI law that may provide clarity for firms under a harmonized, EU-wide regulatory framework.
Although the foundations are technically in force, their implementation is a lengthy process. Most of the law's provisions — including rules for general-purpose AI systems like OpenAI's ChatGPT — is not going to come into effect until 2026 on the earliest, at the tip of a two-year transition period.
“It is crucial for the competitiveness of our companies and the future progress of Europe that both the private and public sectors succeed in developing and using AI in the coming years,” said Caroline Stage Olsen, Denmark’s Minister for Digital Affairs, told CNBC. He called the white paper a “helpful step” towards that goal.
Netcompany's Rogaczewski told CNBC that he presented the concept for a white paper to a few of Denmark's largest banks and insurance firms a number of months ago. He found that while every organization was “experimenting” with AI, institutions lacked a “common standard” to get probably the most out of the technology.
Rogaczewski hopes that the Danish white paper will even be a blueprint for other countries and corporations that wish to simplify compliance with the EU AI law.
Of particular note is Microsoft's decision to affix the rules. “Microsoft’s involvement was important because generative AI solutions often involve algorithms and global technology,” Rogaczewski said, adding that the tech giant’s involvement underscores how responsible digitalization is feasible across borders.
The US tech giant is a key backer of ChatGPT developer OpenA, which received a valuation of $157 billion this yr. Microsoft also licenses OpenAI's technology to firms through its Azure cloud computing platform.
image credit : www.cnbc.com