A legal framework governing the implementation and supervision of EU regulations concerning the use of artificial intelligence (AI) is currently under public consultation, with the proposal as it stands stipulating hefty fines as high as €35 million in some cases.

Two bills have been drafted for compliance with EU Regulation 2024/1689 which lays down harmonised rules on AI.

Taking the lead is the Office of the Commissioner for Electronic Communications and Postal Regulation (Ocecpr).

The public consultation on the bills will remain open until September 16, on the e-consultation platform. There, organisations and individuals alike can provide feedback.

The Ocecpr is designated as the regulator for all matters AI. It is also designated as the notifying authority and as the single point of contact for implementation of the EU Regulation in Cyprus.

The proposed legislation also envisions the creation of an AI Regulatory Sandbox – a controlled environment where startups and researchers can test innovative AI applications in real-world conditions.

For high-risk AI systems, Ocecpr will liaise with other authorities responsible for their respective sectors. High-risk AI systems are defined as applications that pose significant threats to people’s health, safety, fundamental rights or economic wellbeing.

For example, the market supervisor will be the Personal Data Protection Commissioner.

The Central Bank of Cyprus is designated as the supervising authority for high-risk AI systems used by financial institutions.

In its capacity as the Notifying Authority, Ocecpr will be responsible for evaluation, notification and compliance. It will be able to assign evaluations and monitoring to the Organisation for the Promotion of Quality.

Interested organisations will file an application along with a description of their proposed activities and AI systems to be deployed. The Organisation for the Promotion of Quality will provide accreditation.

In exercising its functions, Ocecpr may request technical, financial or any other information from any individual or company. The information will be deemed confidential.

Authorised officials would be able to gain entry into business premises – private homes are exempt – and carry out inspections and gather information about potential violations. They would also have the authority to summon witnesses and interested parties for verbal or written statements.

Obstructing an official during the exercise of their duties will be a criminal offence. If found guilty, a person is liable to six months imprisonment, a fine of up to €10,000, or both.

Administrative sanctions include a written warning, a recommendation to cease the violation, and withdrawal and recall of an AI system from the market.

For minor offences, authorities may impose a fine of up to €30,000, and then €60,000 for repeat offences.

For serious offences, regarding proscribed AI practices, fines on companies can go as high as €35 million or up to seven per cent of global turnover for the previous financial year, whichever number is greater.

Violations by providers, authorised dealers, importers and distributors of AI systems can carry fines of up to €15 million or up to three per cent of global turnover.

Under the proposed legislation, any person or legal entity may file a complaint about a suspected violation. In addition, anyone who considers that they have suffered damage from a decision by the authorities may take recourse at the administrative court.

Cyprus recently unveiled its National AI Strategy 2032, setting out an ambitious long-term roadmap intended to transform public administration, strengthen economic competitiveness and position the country as a trusted regional hub for artificial intelligence.

The document places Cyprus’ ambitions against a rapidly expanding global market, citing international forecasts that expect AI to grow from a market worth $189 billion in 2023 to $4.8 trillion by 2033, while highlighting expectations that AI and automation will reshape businesses across almost every sector during the coming decade.

Earlier this week, economists at the European Central Bank said in an analysis that heady valuations in AI stocks are primed for a sharp market correction, warning that Europe may face a financial crisis if the AI bubble bursts.