Access Control stories
Security teams can now apply one policy model across more AI agents as Bedrock Data adds Google Vertex AI to ArgusAI.
AI-driven attacks are pushing firms to hide systems from the public internet rather than rely on patching flaws after discovery.
New guidance aims to help firms curb data leakage and rogue actions as AI agents and models are embedded in daily operations.
Organisations can now block unsanctioned AI tools and limit agent movement across networks as security teams face rising shadow AI and compliance pressure.
Businesses in Australia and New Zealand are seeking simpler tools as AI adoption and quantum risk sharpen demand for data security.
AI tools have surfaced customer records and other sensitive files at 29% of firms, highlighting weak Microsoft 365 governance.
The appointments come as businesses shift AI from trials to daily operations, putting far greater pressure on data privacy and security teams.
Gaps in visibility are leaving firms exposed, with most finding hidden AI agents in their systems and many suffering incidents.
Despite widespread confidence in governance, UK companies are already seeing AI tools surface sensitive data as Copilot rollouts accelerate.
Machine learning teams could cut drift and operational overhead as Redis folds feature definition, orchestration and serving into one managed platform.
As AI agents spread across workplaces, static credentials are proving too risky for sensitive tasks and customer-facing systems.
Enterprises will need cryptographic proof of AI behaviour, as regulators and customers demand traceability over blind trust.
Growth at Gallagher Security has been driven by culture and long-term investment, with revenue and scale more than doubling since 2021.
The recognition could help New Zealand businesses seeking simpler hybrid-work security, after Nextro completed several Fortinet SASE deployments.
Research teams could see faster target discovery as OpenAI opens GPT-Rosalind to qualified US customers for biology and drug discovery work.
Companies adopting foundation models are being urged to rethink defences as Protegrity’s new tool aims to shield sensitive data during inferencing.
Most firms expect autonomous tools to outstrip guardrails within a year, leaving agent actions hard to see, control and roll back.
A flaw in a Microsoft GitHub workflow could let attackers run unauthorised code and steal repository secrets, Tenable said.
The front-of-house security model now spans 27 countries and 89 clients, reflecting rising demand for blended reception and security roles.
UK regulators are racing to assess whether Anthropic’s Mythos model could speed up attacks on banks and unsettle financial stability.