KS1 (Age 5-7)KS2 (Age 7-11)11 Plus (Age 7-11)KS3 (Age 11-14)GCSE (Age 14-17)PSHE A.I.ESL Spanish Specialist Games

Need help with 28 September 2025 - AI News Quiz for Children Aged 14-17 Years Old? Ask our AI Tutor!

AI Tutor - Lucy

Connecting with Tutor...

Please wait while we establish connection

Hi! I'm Lucy, your AI tutor. How can I help you with 28 September 2025 - AI News Quiz for Children Aged 14-17 Years Old today?

now

Close-up of a gold chess king on a black-and-white board with silver pieces blurred in the background, suggesting strategy, competition, and a decisive endgame.

Need help with this topic? Ask our AI Tutor!

Final boss unlocked: win the board, then explain the physics.

28 September 2025 - AI News Quiz for Children Aged 14-17 Years Old

Welcome to the AI News Quiz for ages 14–17. We’ve turned the week’s biggest stories—about breakthroughs, safety rules, and the people leading AI—into clear, bite-sized questions. You don’t need to be an expert; just bring curiosity and a sense of humour. The aim is to help you understand what matters in AI and why it affects real life.

Choose the answer you think is best and, if you’re unsure, guess! You’ll be told straight away whether you were right and you’ll get a short explanation so you learn something new every time. From smarter robots to ‘red lines’ for safety—and even a cheeky “final boss” idea for testing super-smart AI—this quiz is a fun way to keep up with the news. Ready to play?

1 .

Who told the UN Security Council that the UK’s mission is to “embed AI safely in society”?

Keir Starmer

David Lammy, the UK Foreign Secretary

Peter Kyle, Science Secretary

Yoshua Bengio

At the UN Security Council, UK Foreign Secretary David Lammy set out the UK’s stance on AI: safety, legality and ethics first, with the UK’s AI Safety Institute and the International AI Safety Report cited as core pillars. The message: powerful AI must strengthen peace and security, not undermine it.

2 .

Which AI pioneer added his name to a “Global Call for AI Red Lines” asking countries to set clear no-go zones for risky AI?

Demis Hassabis

Yann LeCun

Geoffrey Hinton

Fei-Fei Li

During UN week, more than 200 leaders signed a public call urging governments to agree global “red lines” for AI - banning things like AI that impersonates people or can self-replicate. Geoffrey Hinton, often called a “godfather of AI,” is among the signatories, highlighting broad, cross-border support for enforceable safety rules by 2026.

3 .

At a big tech showcase, what went wrong with Mark Zuckerberg’s AI smart-glasses demo?

A cooking-help demo glitched and a live WhatsApp video call wouldn’t connect

The AI tried to order 99 pizzas for “research purposes”

The neural wristband kept typing “LOL” on its own

The glasses started livestreaming cat filters on everyone

During Meta Connect, two demos bombed: “Live AI” on Ray-Ban glasses accidentally triggered on many devices at once (a self-inflicted DDoS), and later a WhatsApp video call to Zuckerberg failed due to a rare display-sleep bug. Meta’s CTO explained the issues after the keynote. Classic live-tech oops, not necessarily a product killer.

4 .

Which new UK body was launched to help write a clear “rulebook” for medical AI?

National Commission into the Regulation of AI in Healthcare

The Royal Society of Robot Nurses

The NHS Supercomputing Fan Club

The Big Friendly Data Squad

The MHRA announced a new National Commission into the Regulation of AI in Healthcare to advise on testing, safety and a refreshed regulatory framework due in 2026. The group brings together clinicians, regulators and AI experts so promising tools reach patients sooner - and risky ones are stopped.

5 .

What major upgrade did Google DeepMind roll out so robots don’t just guess and grab?

Teach them to moonwalk while mopping corridors

Give them fabulous wigs so hard hats fit better

Let them plan multi-step jobs by searching the web first

Make them solar-powered by wearing sunglasses indoors

DeepMind released Gemini Robotics-ER 1.5 (the planner) with Gemini Robotics 1.5 (the doer). ER 1.5 can use tools like Google Search to gather rules (e.g., recycling or weather), draft a plan, then hand it to the action model to execute - so robots think before they move. Demos showed tasks like laundry sorting and packing.

6 .

What programme did UK Research & Innovation back to turn piles of studies into quick, trusted advice for policymakers?

A national survey of memes to guide every decision

A gossip-bot for parliamentary corridors

A new rule to replace footnotes with emojis only

An £11.5m AI “evidence synthesis” platform led by Queen’s University Belfast

UKRI funded an £11.5m AI-driven evidence-synthesis programme to scan huge volumes of research and produce reliable summaries for decision-makers. Queen’s University Belfast leads the project, with UK partners building tools to get rigorous evidence to ministers fast. It’s basically a super-librarian for policy.

7 .

How is the UK government using AI to stop fraudsters before they strike?

With an AI tool that scans new policies for loopholes and flags risks early

By replacing every spreadsheet with emoji-only forms

By hiring parrots to shout “Don’t do fraud!” in offices

By making all passwords at least 97 characters long

A Cabinet Office update said the UK blocked over £480m in fraud in a year and unveiled an AI Fraud Risk Assessment Accelerator. It analyses fresh policies for vulnerabilities so fixes happen before scammers exploit them. Trade coverage adds the tool could cut risk-spotting time by up to 80% and may be licensed abroad.

8 .

What huge commitment did NVIDIA make to boost OpenAI’s computing power?

Give every student a free gaming PC for revision

Launch a mind-reading keyboard for coders

Invest up to $100 billion and help deploy at least 10 GW of NVIDIA systems

Build a robot that brings tea to data centres

NVIDIA and OpenAI announced a partnership: NVIDIA intends to invest up to $100 billion and support the rollout of at least 10 gigawatts of NVIDIA-powered data-centre systems for OpenAI, with the first gigawatt targeted for the second half of 2026. That’s a massive scale-up for training and running next-gen models.

9 .

Sam Altman floated a cheeky “final boss” for AI. What would count as truly human-level smart?

Beat every human at chess while cooking pasta

Pass every school exam without revising

Write a hit song in every language at once

Solve quantum gravity and clearly explain the steps

At a Berlin event, OpenAI’s Sam Altman, joined by physicist David Deutsch, suggested a bold AGI test: if an AI genuinely solved quantum gravity and showed its reasoning, we should call that human-level intelligence. It’s half playful, half serious - and sets a very high scientific bar for what “real thinking” means.

10 .

In England’s stroke units, what job is a new NHS AI doing in just seconds?

Choosing the most appropriate waiting-room TV channel

Ordering ambulance snacks automatically for paramedics

Analysing CT brain scans to speed treatment decisions

Writing discharge letters in pirate speak to cheer patients up

Coverage this week confirms every stroke centre in England uses an AI that reads CT scans within seconds, helping doctors choose the right treatment faster. Quicker calls can mean less brain damage and better recovery outcomes across the NHS.

Author: Tara Kemp