Deepseek stories
Agentic AI, zero-day surge, sovereign cloud, and humanoid robots will define IT strategy in 2027, Info-Tech Research Group warns.
Companies may be exposing sensitive data as staff use personal AI accounts for work nearly two-thirds of the time, researchers found.
Most workers are blurring the line between corporate and personal AI use, leaving employers blind to sensitive data shared outside approved accounts.
Workplace AI use is rising faster than company oversight, with a small minority of staff driving most activity and security risks.
Australian startups will get direct access to Chinese tech giants, with a Zhejiang trade mission including Alibaba, Unitree Robotics and Geely.
Sensitive prompts and documents will stay out of model training as ExpressVPN enters AI software with an enclave-based service for Pro subscribers.
New thin-and-light models from Acer to MSI promise local AI and up to 77% faster gaming than Lunar Lake, Intel says.
CrowdStrike rolls out AI security controls for Falcon and deepens Next-Gen SIEM integration by ingesting Microsoft Defender telemetry.
CrowdStrike beefs up Falcon with new AI security controls and links its Next-Gen SIEM to Microsoft Defender to tame sprawling AI estates.
AMD shifts its AI pitch from raw silicon to open software and cloud access as it targets developers and a share of looming trillions.
Red Hat launches AI Enterprise “metal-to-agent” platform and AI 3.3 update to standardise governed AI operations across hybrid clouds.
Anthropic alleges Chinese labs DeepSeek, Moonshot and MiniMax ran vast illicit campaigns to copy Claude and bypass US export curbs.
Multipole AI has launched AIMSOO, a free platform to help SMEs boost visibility in AI-driven search and chat interfaces.
AI is turning romance scams into industrial-scale fraud, with deepfakes and chatbots driving billions in losses, Tenable warns.
AI-powered fraud rings industrialise online romance scams, fuelling USD $5.7 billion in losses and ushering in a 'dark age' for victims.
Agentic AI is set to flood infrastructure with billions of virtual users, forcing nations like Australia to rethink open, heterogeneous compute.
AI in 2026 will bring harsher harms and huge gains, forcing business to prioritise governance, greener models and pragmatic industrial uses.
Voluntary model reviews may leave gaps as advanced AI systems move closer to critical infrastructure and enterprise data.
Current frontier models still fall short of stand-alone cyber defence, with the top performer spotting only 46% of attack evidence in Simbian’s test.
The grant lets the London startup train an air-gapped coding model on UK infrastructure, bolstering supply for defence and other sensitive sectors.
