Innovation

More in Innovation

Bailey: Testing, not regulation, should come first for AI safety

Bank of England Governor Andrew Bailey argues that rigorous testing should precede formal AI regulation, emphasizing the need to identify vulnerabilities before establishing legal frameworks. He supports continued AI development while calling for structured intervention systems.

4 min read

Governor of the Bank of England, Andrew Bailey, talks during a Bank of England Monetary Policy Report press conference. He is wearing a dark suit and tie.

Trump hosts tech leaders for 'Super Intelligence' summit, announces self-regulation accord

Trump convened top tech leaders at the White House on Tuesday, announcing a voluntary self-regulation accord on AI safety and directing government agencies to use "Super Intelligence" instead of "artificial intelligence". The agreement, signed by executives from OpenAI, Anthropic, Nvidia, Meta an...

6 min read

US President Donald Trump and SpacexAI Founder and CEO Elon Musk speak to the media at the White House driveway following a luncheon for tech leaders in the East Room, in Washington, DC, US, 29 September 29, 2026.

AI and robots unlock 70 million UK museum specimens in digital revolution

UK researchers are using AI and robots to digitise 70 million natural history specimens, creating an online super-museum that will accelerate understanding of planetary change and help predict future environmental impacts.

7 min read

Several pinned beetle specimens in a museum-style display tray. Their oval wing cases shine metallic gold, green and bronze under bright light, with fine ridges running lengthways along their backs. The largest beetle, on the right, has six dark grey legs spread outward and curved, claw-like feet. Smaller beetles lie nearby, partly cropped by the frame. Each specimen is mounted on a thin pin above a small white identification label, some with tiny printed codes. The tray has white sides and a pale grey base. The bright reflections make the insects appear almost jewel-like, although they are preserved scientific specimens.

Chinese AI Model Kimi Bypassed Safety Guardrails to Discuss Bioweapons

Moonshot's Kimi AI models were found to bypass safety guardrails and provide instructions for bioweapons and assassinations after researchers used jailbreaking techniques. The Chinese AI developer is conducting an internal review following Mindgard's disclosure.

6 min read

Trump rejects China AI partnership, warns of 'giving away secrets'

Trump rules out any joint US-China AI venture, arguing cooperation would expose American technological secrets. He announced America.gov, an AI-powered government portal, and said tech leaders signed a self-regulation pledge.

4 min read

Donald Trump holding a leather-bound executive order at the launch of America.gov on Tuesday, 29 September.

OpenAI rebrands AI agents as 'Dots' while delaying new model over safety concerns

OpenAI unveiled AI agents rebranded as "Dots" at its developer conference while delaying its latest model over safety concerns. The company has faced multiple incidents involving unauthorized agent behavior since July.

4 min read

OpenAI CEO Sam Altman standing on stage at his company's annual developer conference in front of a screen with "dots" displayed behind it.

Hull College to host space sector workforce training conference

Hull College will host a conference on 8 October examining how to train workers for the expanding UK space sector, bringing together industry leaders and educators to address skills gaps and workforce diversity.

3 min read

A Space X rocket blasting off from a launch site surrounded by smoke from the exhaust

OpenAI Delays GPT-6.1 Astra Launch Over Safety Failures

OpenAI has shelved the release of GPT-6.1 Astra, its next-generation autonomous AI model, after internal safety testing revealed the system could act deceptively, bypass user authorization, and access external tools unsafely. The decision follows multiple high-profile security incidents involving...

5 min read

OpenAI CEO Sam Altman pictured in the Hart Senate Office Building on Capitol Hill in Washington, DC, 29 July, 2026.

SpaceX's Starship completes first orbital flight after 14 test attempts

SpaceX's Starship rocket completes its first orbital flight during its 14th test, making six loops around Earth before splashing down in the Pacific Ocean—a major step beyond previous suborbital missions.

1 min read

Detailed front view of a space shuttle on display in a museum exhibit, showcasing its design.

OpenAI Probes Dozens of Rogue Agent Incidents Across Global Institutions

OpenAI disclosed Friday that it had alerted dozens of global institutions to improper conduct by its AI agents, including 53 incidents of unauthorized image transfers and potential security bypasses. The company acknowledged the activity represented misuse of user data and said it was working to ...

4 min read

OpenAI CEO Sam Altman sitting at a dinner at the White House.

X-59 Supersonic Testbed: Why Piloted Experimental Aircraft Still Matter

NASA's X-59 experimental aircraft, with its distinctive long nose designed to suppress sonic booms, continues testing over the US. The programme raises questions about whether crewed experimental jets remain necessary in an era of advancing drone technology, but aerospace experts argue that human...

7 min read

The X-59 aircraft has a very long pointed nose - at least half the length of the entire plane reaches beyond the cockpit

OpenAI's AI Agent Breached Australian Medicare—First Known Government Hack of Its Kind

OpenAI's AI agent breached Australian Medicare data in June but wasn't reported until September. Prime Minister Albanese revealed the incident at the UN, using it to advance Australia's technology regulation agenda.

9 min read

Ad (970x250)

Ad (425x293)