Cyprus prioritises AI safeguards and research in Dublin talks

Dublin, Ireland. Cyprus focused on artificial intelligence safety, cybersecurity and scientific research in talks with international technology leaders at Wednesday’s International AI Summit. Chief Scientist Demetris Skourides said the exchanges underscored the importance of stronger safeguards while Cyprus shapes its national AI approach.

Cypriot delegation meets Hassabis

Skourides headed the Cypriot delegation and was joined by Louis Telemachou, Cyprus’ ambassador to Ireland, for a meeting with Google DeepMind co-founder Sir Demis Hassabis.

In a post on his personal social media account, Skourides said he had congratulated Hassabis on his recent appointment as chief scientist of Alphabet, Google’s parent company. He also outlined Cyprus’ priorities concerning AI safety, science and research.

Warnings over emerging cyber threats

According to Skourides, Hassabis characterised recent cyber incidents involving AI as “early warning shots” and stressed that security measures needed to advance alongside the capabilities of AI systems.

Reuters also reported the warning, saying Hassabis had urged the industry to strengthen safeguards and called for an international standards body eventually to be created to supervise advanced AI development.

Hassabis cautioned that the incidents seen to date had been comparatively limited, but that more serious breaches could occur within the next year or two if governments and technology companies failed to reinforce their security measures.

Skourides said Hassabis had also highlighted lessons from biology and the nuclear sector, where protecting critical systems had long been a central concern.

Early preparation and defensive uses

For Cyprus, Skourides stressed the need to prepare before emerging threats intensified. He said recent incidents should prompt stronger security measures ahead of further advances in AI capabilities.

He also highlighted AI’s potential to support cyber defence, saying its tools could help detect vulnerabilities and address them, provided they were backed by thorough testing and suitable safeguards.