Why Cyprus AI Monitor exists

Public discussion of AI in Cyprus often relies on announcements, plans and estimates. They have their uses, but they rarely say what was measured, by whom and for which period. Cyprus AI Monitor collects the figures that can be verified and shows each one with its source, its quality tier and the date it was last reviewed. Where no reliable figure exists, we say so instead of filling the gap with an estimate.

The site publishes 14 metrics in 7 pillars, and each one links to the document its figure was taken from. All 14 metrics →

Cyprus AI Monitor is not a government publication and is not part of the national AI strategy. A national strategy is a policy document: it sets objectives and describes the actions the government intends to take to reach them. The monitor reports the current position, as shown by published evidence. It can be used to follow progress against a strategy, but it does not assess the strategy itself.

Two strategy documents are relevant. We last checked the status of both on 14 Sept 2026.

This section describes where the work is heading, not what the site does today. The aim is for more of the picture to come directly from the systems where the evidence is recorded, and for every score to come from a procedure that anyone can check.

Today, 8 of the 14 metrics update automatically from their publisher, and the rest are compiled by hand from documents. None of the steps below is in place yet. Each will be described on the method page before it changes a published figure.

• Data read from its sourceConnections to the systems that hold the evidence, such as government databases and public registers, so that a figure is read automatically from where it is recorded rather than copied from a report. Each connection would log what it read and when, so the figure can still be checked against its origin.
• Measures of what AI changes in practiceMost published figures describe capacity, such as skills, infrastructure and adoption. Operational data would make it possible to measure results as well: for example, how long a public service takes to handle a request before and after it starts using AI. Each such measure would state how the comparison was made, so that a change is not credited to AI when something else explains it.
• Why one number is not enoughA country can look strong on one measure and weak on another. Many businesses may use AI while the economy gains little from it, and a ranking on either measure alone would miss that. We intend to show the measures side by side and ask a simple question of each country: does another country do better on all of them at once? If one does, that is where improvement is clearly possible. If none does, the country is among the best results so far, a group economists call the Pareto front, and doing better on one measure means deciding what matters more. The examples below show why this matters, and the figure after them shows the idea with invented values.
• Scoring that adapts and can be reproducedThe scoring will change as sources and measures are added, but never silently. Each change will be published as a new version of the method, and each score will name the version that produced it, so an earlier score can be recalculated exactly. Scores will be calculated by a published procedure from published inputs. Each procedure will be checked against worked examples before it is used, and anyone who runs it on the same inputs will get the same result.
• Reported experience, kept separateFeedback from people who use AI-assisted services may be added as a further objective. If it is, it will be collected by a stated method, labelled as reported experience and shown separately from measured figures. A voluntary survey will not be presented as a national rate.

Cyprus AI Monitor is an independent initiative and welcomes contributors. The picture it gives is only as complete as the evidence behind it, and several parts are still missing. Accurate data that can be verified is the contribution that helps most.