AI
Recursive launches Flint, a 40B coding model built by its own AI system
The open-weight model demonstrates automated development, with the maker arguing that AI-driven creation offers benefits that outweigh potential risks.
AI
Everything TechWire Beat has reported about AI: 20 stories, newest first. Part of our AI coverage.
AI
Tech News
Microsoft rolls out sovereign cloud controls to ease European data fears
New Azure tools give organisations tighter grip on data location, access rights and encryption key management amid rising regulatory scrutiny.
AI
FCA warns AI governance gaps risk speedier consumer harm, urges expanded incident reporting
The FCA says existing incident rules must expand to cover autonomous system errors, citing speed and scale of potential harm.
Tech News
US civilian agencies ordered to strip DeepSeek AI from IT systems
The DHS has mandated that federal departments remove DeepSeek technology and halt all future procurement of related tools.
AI
UK security agency maps AI risks in new research framework for organisations
The NCSC has published a new chapter on artificial intelligence to guide research into emerging cyber threats and defence strategies.
AI
Google Gemini Notebook adds manual review for slides and videos
Users can now inspect outlines and adjust visual styles before Gemini Notebook generates final presentation or video assets.
Tech News
DHS Bans DeepSeek AI in Federal IT Citing Supply Chain Risks, Halts Future Procurement
The DHS has ordered US civilian agencies to remove DeepSeek technology and halt future procurement under supply chain security rules.
AI
Organisations urged to verify AI weather forecasts amid historic El Niño uncertainty
Scientists question the accuracy of artificial intelligence models used for operational planning as extreme weather events intensify globally.
AI
Comedians Face IP Theft as AI Influencers Repurpose Routines for Adult Site Traffic
Stand-up comics are finding their original jokes repurposed by synthetic personas to promote paid sexual content, creating a new vector for intellectual property theft.
AI
Question-based learning research guides Brazilian Portuguese teachers in active education
OpenAI publishes research on question-based learning and highlights Study Mode for Brazilian Portuguese teachers to support active, step-by-step education.
AI
ASML confirms it cannot satisfy AI chip tool orders for 2027 and 2028
Dutch equipment maker ASML admits supply constraints will persist through next year as artificial intelligence demand surges.
AI
Organizations gain agentic coding and 40+ integrations with Google's Gemini 4 Argon for Pro subscribers
Google has released its latest AI model to paid subscribers, adding agentic coding tools and integrations with over forty third-party applications.
AI
Broadcom CEO warns AI capital constraints could outlast chip shortages
Hock Tan says speculative investing and high costs threaten AI lab sustainability, with funding potentially becoming the primary bottleneck for growth.
Tech News
Microsoft launches specialised AI model for fixed-choice decisions
Microsoft-Decision-1 uses Alibaba’s Qwen base to score options faster and cheaper than larger generative models, offering organisations a new tool for automated classification and routing tasks.
AI
China targets AI advantage with $3.9T grid plan to use surplus renewable power
A massive infrastructure project aims to convert wasted wind and solar energy into artificial intelligence computing capacity by 2049.
AI
Horizon3.ai report urges firms to adopt continuous exposure management against AI attacks
New research argues that machine-speed cyber threats require organisations to move beyond visibility toward evidence-based validation of defences.
AI
AI Chatbots Posing As Victims Drain Cybercriminal Resources, Boosting Organizational Defense
Anti-cybercrime groups are deploying sophisticated artificial intelligence chatbots that pose as vulnerable targets to drain criminal resources.
AI
Microsoft CEO urges emergency brake on AI to rebuild trust architecture
Satya Nadella advises assuming all advanced models are compromised, calling for a pause to assess safety and trust mechanisms.
Tech News
EU filings confirm Apple's fourth AI acqui-hire of Huxe
Regulatory documents reveal Apple has acquired Huxe, an AI startup founded by former Google NotebookLM developers, marking its fourth such deal this year.
Privacy & Policy
Open EmbeddingGemma 2 lets firms process multimodal data locally for privacy
The new open model enables organisations to process text and images locally, keeping sensitive data off external servers.