AI Safety stories
The update could let companies use a single security key to approve payments, documents and AI actions without custom cryptography.
Businesses using AI coding tools face repeatable security gaps, as the new index found an average 15 vulnerabilities in each codebase.
Newsrooms, designers and robotics teams are set to gain new AI tools as NVIDIA expands detection, agent and edge software at SIGGRAPH.
Enterprise teams can now switch on controls for AI agents and APIs faster, with Salt Security bundling 100 pre-built policies into its Policy Hub.
Limited internal access resumed after monitored tests found the model could bypass safeguards, prompting OpenAI to tighten trajectory-level monitoring.
The recognition reflects rising demand for safer, production-ready AI in robotics and industry, where testing and reliability now matter most.
Shoppers in the UK and Europe face a growing risk of cloned retail sites as Gen blocked 114.2 million fake e-shop attacks in six months.
The platform lets security teams run AI pentests without exposing customer infrastructure data to external models.
Generative AI is helping Tehran speed up phishing, malware and influence campaigns, while making attacks harder to attribute.
Teens using ChatGPT will face tighter safeguards and parental oversight as OpenAI expands study tools and limits harmful content.
Developers can now use Google Cloud's examples to build and govern Gemini-powered agents for approvals, compliance and data workflows.
Enterprises are turning to AI agents that can handle complex service requests, raising the stakes for vendors competing on reliability and governance.
A blind employment pathway he founded has helped lift job prospects for graduates, as Professor Iain Murray won Western Australia's top tech honour.
A free phone service aims to help Australian small businesses move beyond early AI trials, after research found most do not know where to begin.
With AI-related defects threatening user trust, the new platform pairs automated test creation with human checks across global devices and markets.
Businesses that treat AI as a mere efficiency tool risk missing bigger gains from redesigning workflows, according to One NZ's Chief Executive Officer.
Only 22% of Australian organisations say they are ready to govern AI safely, as spending and use surge across daily operations.
Trust remains a hurdle as younger Australians increasingly use AI to research investments, superannuation and retirement planning, a survey shows.
AI projects that remain stuck in pilot stages may miss the shift to agentic systems able to manage tasks with limited human oversight.
Only four in 10 Singapore professionals say they can spot AI misinformation, prompting a year-long reading drive for better judgement.
