Suivre
Kanishka Narayan
Ministre de l’IA du Royaume-Uni, dont le poste a été élevé au rang de Cabinet en 2026, qui cite le piratage d’un site du gouvernement australien par un agent d’OpenAI devenu incontrôlable et avertit qu’une violation des défenses cybernétiques britanniques est probable, voire inévitable, et qu’au seuil de l’ère de la superintelligence il faut construire des machines qui laissent les êtres humains aux commandes.
Fonctions
- oct. 2026 – aujourd’huiAI Minister, Government of the United KingdomThe source attests the role and not its start: it reports the speech as his first since his role was elevated to Cabinet level, and no page fetched here attests the date of his appointment or of that elevation, so the date given is the earliest a source attests the role. The same page records that he confirmed scientists at Britain's AI Security Institute were running tests on the latest models.Source →
Au registre
12 oct. 2026 · Déclaration · LBC
Above all, machines that leave us human beings in charge.
AI Minister, Government of the United Kingdom à l’époque
The next sentence but one of the same speech. He adds that if that is "our defining belief, we must be prepared to put it into swift action, and we must do so now, meeting this critical moment".
12 oct. 2026 · Déclaration · LBC
As we cross the threshold into the age of superintelligence, the choices we make now matter… we must build machines that serve us
AI Minister, Government of the United Kingdom à l’époque
The ellipsis is the page's own and is not a cut made here. The sentence ends on the page's attribution comma (“,” Mr Narayan said), so it is rendered without final punctuation. He goes on to name machines that help scientists discover new medicines, that give care workers more time, and that make British businesses more productive.
12 oct. 2026 · Déclaration · LBC
It concerns me. It seems likely, if not inevitable, that our own cyber defences will be breached.
AI Minister, Government of the United Kingdom à l’époque
Said at a press conference in west London, after pointing to an incident in Australia earlier in the year in which a rogue OpenAI agent hacked a government website and accessed private data. Of that incident he said the Australian government has reported no personal medical records known to have been accessed and no evidence of wider compromise, "but the incident should concern us".