AI Safety stories
It aims to cut the steps IT administrators take by routing natural-language requests to specialist workflow agents while keeping human approval for changes.
Breaches are rising as AI tools widen access inside firms, leaving security teams struggling to spot data flows that Zero Trust was built to limit.
Regulated teams can now keep AI processing inside GitLab Dedicated, with new secret handling, spending caps and security fixes added in 19.3.
Startups risk leaked keys, quota throttling and surprise bills unless they tighten controls as AI prototypes move into production.
The deal extends Fortinet's reach into AI runtime protection as companies race to secure agents, models and tools across production systems.
Australia's new scams rules are adding pressure on firms to detect fraud faster as AI-powered attacks expose weak governance and identity controls.
Misconfigured models are giving attackers a fresh route into cloud systems, raising the risk of data theft and service compromise.
The update aims to help campaigns and publishers prove authentic footage as deepfakes and synthetic media make video origins harder to verify.
Enterprise security teams now have a proposed benchmark for limiting AI agents to approved actions inside live systems.
Most firms are still struggling to verify the data behind AI agents, even as more move them from pilots into production.
Governance gaps are widening as 60% of CIOs plan AI agent investment within 12 months, a survey of 1,000 leaders found.
Researchers can claim up to USD $1 million for breaking Vercel Sandbox's isolation, as the cloud provider opens its boundary to public scrutiny.
Labour-heavy services face the biggest squeeze as AI shifts spending towards infrastructure, data management and security, Forrester says.
IT teams in Australia and New Zealand can now police AI apps on managed Macs as Jamf adds endpoint-level controls and audit reporting.
Bias in AI is becoming a business risk as researchers warn English-language models still miss cultural nuance in non-Western markets.
Customers can now get faster help in more than 50 languages as NAVEX brings its Nira assistant into support and case creation.
Customers can query live pricing data in Microsoft Copilot, Claude or ChatGPT as Pricefx opens its platform to approved enterprise AI assistants.
Enterprise software teams can switch large language models without changing release checks, reducing lock-in as AI coding outpaces testing.
Enterprise buyers are weighing governance and auditability more heavily as the AI vendor secures top marks across five analyst categories.
The hire signals Tokenovate's push to make its post-trade automation platform easier for institutions to deploy at scale.
.webp)
