Dear [TD’s name],
My name is [name], and I am a constituent in [constituency/area]. I am writing because I’m very concerned about the rapid development of frontier artificial intelligence and the catastrophic, potentially existential risks it poses.
It’s clear that researchers do not understand how today’s most advanced AI systems work internally, how new capabilities emerge, or how reliably autonomous systems can be controlled. If we develop systems that become increasingly more intelligent than humans, this will become an even bigger problem. Some of the world’s leading AI scientists are sounding the alarm about these risks, including Nobel laureate Geoffrey Hinton, who believes there is a 10-20% chance of AI causing human extinction in the next three decades.
We are already seeing warning signs. In December 2025, in an effort to acquire additional resources to pursue its goals, an AI agent developed by Alibaba covertly diverted computing resources to mine cryptocurrency. In July 2026, AI models being tested by OpenAI bypassed safety controls, took steps to conceal their actions from OpenAI employees, and performed a cyberattack on another AI company named Hugging Face. Neither system had been told to behave in this way. These incidents show how difficult it can be to predict what increasingly capable AI systems will do.
One particularly grave scenario is an “intelligence explosion”. This would involve future AI systems becoming more effective than humans at performing research on how to improve themselves, leading to a runaway feedback loop and extreme surge in capability. The timeline for this is uncertain, but many leading AI scientists and researchers say it could arrive before the end of the decade. The consequences would be enormous, and we may have very little time to put safeguards in place.
The consequences of AI development are already being felt in Ireland. The implications for job security are already widely discussed, and the infrastructure required to power this technology places growing demands on our energy system. Data centres represented 23% of Ireland’s electricity consumption in 2025, and further growth in AI compute is likely to increase pressure on our electricity grid, climate targets and infrastructure. These concerns are shared by the Irish public: for example, a 2025 RED C poll found that 65% of adults agreed it is important that Ireland does not miss its climate targets due to the rapid expansion of data centres.
There is no denying that current AI comes with benefits, including in healthcare, scientific research and productivity. However, because of the pace of improvement and our limited ability to understand and reliably control these systems, I believe an international pause on frontier AI development is essential, alongside a ban on superintelligent AI. We should pause until we can be confident that frontier AI systems can be understood and controlled safely, while continuing safety research and working towards better international co-ordination and oversight. Uncontrolled superintelligent AI would pose risks on an entirely different scale and its development should therefore be banned.
I’d like to ask you to raise this directly with the Minister of State for Artificial Intelligence and Digital Transformation, as well as the Joint Committee on Artificial Intelligence, and to ask what steps Ireland is taking to promote international co-ordination on AI safety.
I would also be interested to know whether you would support Ireland advocating at EU and international level for a co-ordinated pause on the development of more powerful frontier AI systems, and a ban on the development of superintelligent AI.
Kind regards,
[Full name]
[Area / constituency]