AI Safety stories
It aims to help businesses build AI agents that use live data, keep to strict permissions and avoid hallucinating over filings and news.
KnowBe4 has backed the Open Secure AI Alliance, arguing AI security depends on governing agent behaviour as well as models.
Businesses using AI agents can now approve or block individual commands in real time, reducing the risk of authorised access doing unauthorised things.
Governance features aim to help security teams prove AI controls to boards and regulators as shadow AI and agentic risks spread.
Security teams can now block risky AI prompts in real time as Reco expands its platform from visibility into browser-level enforcement.
Privacy-conscious professionals can now search sensitive files on-device, as Synsira's Kind Local Pro keeps data off cloud servers.
Enterprises that adopt agentic AI early could gain an edge as governance, integration and deployment hurdles keep most firms stuck in pilots.
The new platform aims to cut breach risk as firms let AI agents use more systems, with 40% of developers leaving their access in place.
The new model is designed to halve operating costs for Microsoft's MDASH security system as automated attacks grow cheaper and more frequent.
Fashion retailers can now curb costly AI shopping errors as Algolia adds guardrails tied to live stock, prices and business rules.
European firms face mounting AI compliance pressure as ISACA plans a Munich conference to address governance, cyber resilience and digital trust.
Downstream AI data leaks have more than doubled in a year, with connected services now returning unauthorised information to staff and agents.
Banks and insurers could soon grant AI agents narrower, time-limited access to sensitive systems under Daon's new patent.
Businesses can now query governed data across cloud systems in plain language, as Google Cloud opens its conversational analytics tools more widely.
Despite near-universal uptake, most ANZ SMEs still distrust AI to make cyber security decisions as incidents continue to disrupt work.
Designed to curb data leakage and keep human staff in charge, the framework routes AI outputs by role inside Microsoft systems.
Most Britons think the government is moving too slowly, with 58% saying the UK is not ready for AI's impact on jobs and the economy.
New controls for AI agents and machine identities come as Saviynt passes USD $300 million in annual recurring revenue, with bookings up more than 80%.
Security chiefs face a widening breach risk as most firms plan to use AI agents, yet barely a third feel ready to secure them properly.
A growing share of consumers want businesses to prove AI is powered responsibly, with younger New Zealanders most worried about its environmental costs.
