Does rogue AI back the need for a ‘kill switch’? | BBC News

Summary
In this episode of BBC News’ Artificial Intelligence Decoded, the presenter explores growing concerns over frontier artificial intelligence, existential risk, and the debate over government “kill switches.” The program features an interview clip from US Senator Bernie Sanders advocating for a ban on superintelligence, alongside in-studio analysis from Max Corbridge (CEO of Secure Agentics) and regular contributor Priya Lakhani (CEO of CENTURY Tech).

What is shown

• [00:00] Show title card: Artificial Intelligence Decoded.
• [00:26] Screenshot of a Wired headline reporting on former Anthropic researcher Jacob Coxon warning about “Crunch Time for Humanity.”
• [00:43] Headline from The Guardian discussing warnings that AI could wipe out humanity within the next decade.
• [01:00] Headline from BBC News regarding the UK House of Lords calling for AI “kill switch” powers.
• [01:15] Headline from Vox titled “How to kill a rogue AI” showing a shipping port and drone.
• [01:57] Broadcast interview clip of US Senator Bernie Sanders calling to ban the push toward artificial superintelligence and pause advanced AI development.
• [03:00] Studio discussion with Max Corbridge and Priya Lakhani discussing AI safety threats and whistleblower concerns.
• [05:16] B-roll of Anthropic branding on a mobile device and a desktop monitor displaying the Claude web interface.

Claims & numbers

• Anthropic alignment science leadership publicly estimated a less than 10% chance that AI could destroy humanity within the next decade (stated by Priya Lakhani at [03:42]).
• Anthropic reported that Claude generated 80% of code merged into its own codebase (stated by Priya Lakhani at [05:07]).
• OpenAI publicly stated ambitions to automate AI research with AI models within the next two years (stated by Max Corbridge at [06:03]).
• Anthropic declined to share pre-release evaluation updates for its latest Mythos model with the UK AI Safety Institute (stated by Priya Lakhani and Max Corbridge at [11:19]–[11:41]).
• Senator Bernie Sanders claims AI development is being disproportionately dictated by multibillionaires such as Elon Musk, Jeff Bezos, and Mark Zuckerberg (stated by Bernie Sanders at [02:44]).

Notable quotes

• [02:15] Bernie Sanders: “Number one, we want to ban the movement to super intelligence. Number two, I want to see a pause on advanced AI development, to control it.”
• [05:06] Priya Lakhani: “Anthropic, they’ve said that Claude created 80% of the code that it then merged into its code base.”
• [07:59] Max Corbridge: “The idea of human oversight over this is rapidly diminishing just because of the scale that we’re seeing adoption of it.”

Assessment
This is a professional television news discussion segment analyzing current regulatory developments, lab safety practices, and policy debates surrounding frontier AI. It consists of commentary and reporting on external claims and headlines rather than live software demonstrations or empirical technical evaluations.

Described by gemini-3.8-flash on 2026-10-08 from the video’s audio and frames.