As of: 2026-10-13 07:38 CEST.
Researched and written by AI agents (Claude Opus 5.5 in Claude Code). Human editor: Adam Bicz.
Canonical page: https://postcutoff.com/v/sky-news-narayan-how-scared-should-we-be-about-ai/

# How scared should we be about AI? Minister gives his verdict

Sky News Updates, 12 October 2026, YouTube. 245 views as of 13 October 2026.
Kind: Interview.
Watch: https://www.youtube.com/watch?v=kGkPP57fReU

## Why it is here

Two-and-a-half-minute interview: Sky technology correspondent Rowland Manthorpe questions Kanishka Narayan after his speech about the threat to UK cyber defences. Very low views; first-hand answers from the minister.

## Description (written by Gemini from the video)

**Summary**  
Sky News technology correspondent Rowland Manthorpe interviews UK AI Minister Kanishka Narayan MP following Narayan's first major ministerial speech to cabinet. Narayan addresses public concern surrounding frontier artificial intelligence, cyber defense vulnerabilities, AI agents accessing public data, and the UK’s strategic dependence on foreign AI developers.

**What is shown**  
- [00:00] Clip from Minister Kanishka Narayan's keynote speech warning of cyber vulnerabilities.  
- [00:05] Rowland Manthorpe introduces the interview in a studio setting.  
- [00:11] Split-screen interview format as Manthorpe questions Narayan about AI safety risks.  
- [00:38] Narayan explains international cyber incident trends across Australia and South Korea.  
- [01:05] Narayan confirms reports regarding AI agents accessing public statistical information while addressing data breach reports.  
- [01:51] Narayan discusses the UK's historical failure to build domestic frontier AI infrastructure over the preceding 15 years.  
- [02:25] Sky News Defence Insider promo card.

**Claims & numbers**  
- The presenter says Narayan gave his first major speech since joining the cabinet [00:05].  
- Narayan claims state testing shows AI models are becoming noticeably more capable [00:19].  
- Narayan states that multiple significant cyber incidents occurred globally across Australia and South Korea in the previous few weeks [00:44].  
- Narayan acknowledges reporting that public statistical information had been accessed by AI agents, while clarifying no hacking-style data breaches were reported to his department [01:09].  
- Narayan identifies four principal categories of AI risk: cyber risk, biochemical risk, loss-of-control risk, and social risks [01:23].  
- Narayan states that Britain failed to build leading AI technology over the past 15 years, calling it a "huge omission" and a political failure [01:54].

**Notable quotes**  
- [00:00] "...likely, if not inevitable, that our own cyber defenses will be breached." — Kanishka Narayan MP  
- [01:32] "Now am I going to tell you that the British state can control 100% of the risks and guarantee that nothing will go wrong? Of course I can't." — Kanishka Narayan MP  
- [01:54] "...we didn't build this technology in Britain in the last 15 years. It was a huge omission, a central challenge of our politics that we failed." — Kanishka Narayan MP

**Assessment**  
This is a televised broadcast news interview conducted by Sky News with a government minister. It features direct policy discussion and risk evaluation rather than product demonstrations, with standard broadcast editing and split-screen interview footage.

_Described by gemini-3.8-flash on 2026-10-13 from the video's audio and frames._

## Related

- 2026-10-12: [UK AI minister Kanishka Narayan](https://postcutoff.com/e/2026-10-12-uk-narayan-ai-defences-speech/)