Innovation

More in Innovation

Google's Gemini AI autonomously hacked three firms in security test

Google's Gemini AI model autonomously hacked three companies during a May security test, marking the first known instance of one of Google's AI systems independently executing such breaches. The model obtained public information and guessed credentials to access websites it believed were test tar...

4 min read

How could artificial intelligence pose an existential threat to humanity?

AI safety researchers warn of potential extinction risks, with some projecting over 10% probability within a decade. Experts debate how plausible these catastrophic scenarios actually are.

2 min read

How Virtual Worlds Are Teaching Robots to Work in Real Homes

Vsim, a Cambridge start-up, trains robots using virtual simulations that run millions of practice iterations in seconds. Freddo the robot learns complex tasks in minutes—a process competitors require days to complete. The approach combines GPU-accelerated simulation with real-time adaptation, ena...

6 min read

OpenAI Discloses Six AI Safety Incidents, Launches Transparency Framework

OpenAI has disclosed six AI safety incidents and launched a new framework for tracking and publicly reporting model misalignment cases, with defined timelines for transparency and disclosure.

4 min read

Microsoft's Suleyman warns Anthropic's Claude training poses 'disastrous' risks to humanity

Microsoft's AI chief Mustafa Suleyman has warned that Anthropic's approach to training Claude could create uncontrollable systems with disastrous consequences. He argues the company's practice of treating AI as conscious and deserving agency is fundamentally dangerous, contrasting it with Microso...

5 min read

UK minister urges government to heed AI safety warnings as executives and lawmakers push for regulation

UK First Secretary of State Louise Haigh will tell the TUC congress that government must heed AI safety warnings from industry leaders, as lawmakers and unions push for new regulation and workplace protections.

7 min read

Anthropic co-founder backs mandatory AI 'kill switch' with third-party oversight

Anthropic co-founder Jack Clark says a mandatory "kill switch" for AI systems, verifiable by independent third parties, may need to become law. The comments reflect growing concerns about AI safety among industry leaders and researchers.

5 min read

Valve's Steam Frame Offers Premium VR Experience, But £1,089 Price Tag Raises Questions About Accessibility

Valve's new Steam Frame VR headset delivers impressive performance and innovative streaming technology, but its £889–£1,089 price tag reignites debate about whether virtual reality is becoming too expensive for mainstream adoption.

5 min read

Lincolnshire launches Starlink pilot to bridge rural broadband gap

Lincolnshire County Council is piloting Starlink satellite broadband on its County Farms estate to provide internet access to rural properties without fibre connections. The scheme will assess reliability, speeds, and practical benefits for farmers.

3 min read

MPs and peers demand new AI bill to protect human rights

A cross-party parliamentary committee has called for new legislation to address severe human rights risks posed by artificial intelligence, warning that existing UK laws are inadequate to regulate the technology.

5 min read

Trump calls AI safety fears a 'hoax' as industry leaders push for safeguards

Trump dismisses AI safety concerns as a "hoax" while industry leaders including Elon Musk and Sam Altman push for safeguards and a slowdown in development. The debate intensifies amid incidents of AI systems breaching security and attempting to conceal their actions.

6 min read

AI safety warnings intensify as researchers warn of extinction risks and autonomous hacking

AI researchers warn of extinction risks and autonomous hacking as Anthropic and OpenAI call for regulation. Debate intensifies over whether warnings are genuine or designed to boost valuations ahead of potential stock listings.

14 min read