SecurityBrief New Zealand - Technology news for CISOs & cybersecurity decision-makers
New Zealand

AI Safety Stories - Page 2

Global Stories

AI Safety stories

QNX to unveil AI robotics demos and research report
Robotics

QNX to unveil AI robotics demos and research report

Its research aims to show developers why deterministic software is becoming crucial as AI robots move into shared, safety-critical spaces.

By Sofiah Nichole Salivio • 4 min read • Last week
Proofpoint opens Paris innovation centre for AI security
Data Protection

Proofpoint opens Paris innovation centre for AI security

Growing AI security fears are driving Proofpoint’s European expansion, with the Paris site aimed at helping customers meet local regulatory demands.

By Sofiah Nichole Salivio • 4 min read • Last week
Anthropic hands Petri AI test tool to Meridian Labs
AI Security

Anthropic hands Petri AI test tool to Meridian Labs

The move gives researchers and regulators a more neutral way to probe model deception and harmful behaviour as AI safety scrutiny intensifies.

By Sean Mitchell • 4 min read • Last week
Cognizant launches Secure AI Services for enterprises
Data Protection

Cognizant launches Secure AI Services for enterprises

Enterprises adopting AI in regulated sectors face fresh risks from model tampering and agent misuse, which Cognizant aims to address.

By Mark Tarre • 4 min read • Last week
AWS previews Bedrock AgentCore payments for AI agents
Hyperscale

AWS previews Bedrock AgentCore payments for AI agents

Developers can now let AI agents pay for paid content and services in real time, with US East, US West, Europe and Asia Pacific support.

By Karen Joy Bacudo • 5 min read • Last week
Why trust is the bottleneck for AI-driven operations
DevOps

Why trust is the bottleneck for AI-driven operations

The shift to autonomous IT is stalling because teams will only let AI act when its decisions are transparent, explainable and controlled.

By Karthik SJ • 4 min read • Last week
OpenAI launches GPT-5.5-Cyber for vetted defenders
Firewalls

OpenAI launches GPT-5.5-Cyber for vetted defenders

Vetted security teams will get fewer refusals on authorised tasks as OpenAI tightens access around its most permissive cyber model.

By Mark Tarre • 4 min read • Last week
OpenAI replaces ChatGPT default model with GPT-5.5 Instant
Generative AI

OpenAI replaces ChatGPT default model with GPT-5.5 Instant

Users may notice fewer errors as the chatbot’s default switches to GPT-5.5 Instant, which OpenAI says cuts hallucinations by 52.5%.

By Mark Tarre • 3 min read • Last week
Anthropic adds dreaming & multiagent tools for Claude
DevOps

Anthropic adds dreaming & multiagent tools for Claude

Developers get new ways to boost Claude agents’ accuracy and scale, as Anthropic rolls out memory, grading and parallel task handling.

By Mark Tarre • 4 min read • Last week
OpenAI launches three new voice models for developers
Customer Experience

OpenAI launches three new voice models for developers

Developers could soon build voice apps that handle tasks and translations in real time, as OpenAI adds three new audio models to its API.

By Mark Tarre • 4 min read • Last week
Rapid7 joins OpenAI cyber programme to speed defence
Digital Transformation

Rapid7 joins OpenAI cyber programme to speed defence

The tie-up could help security teams cut false alarms and patch faster as automated attacks shrink defenders’ reaction time.

By Joseph Gabriel Lagonsin • 5 min read • Last week
OpenAI rolls out ChatGPT Trusted Contact for adults
Generative AI

OpenAI rolls out ChatGPT Trusted Contact for adults

Adults using ChatGPT can now name a trusted contact, giving OpenAI a new way to alert someone in serious self-harm cases.

By Mark Tarre • 4 min read • Last week
Malicious OpenClaw skill spreads Remcos RAT & GhostLoader
Crypto

Malicious OpenClaw skill spreads Remcos RAT & GhostLoader

AI agent workflows are being targeted by a fake OpenClaw skill that installs Remcos RAT and GhostLoader on Windows, macOS and Linux.

By Sofiah Nichole Salivio • 4 min read • Last week
AI inference becomes core operational workload in firms
DevOps

AI inference becomes core operational workload in firms

Most firms are now running AI in production, with hybrid clouds and security controls becoming crucial as inference overtakes training.

By Sofiah Nichole Salivio • 4 min read • Last week

Stories from Other Regions

AI Safety stories

Persistent, IIT Kharagpur launch AI innovation lab
EduTech

Persistent, IIT Kharagpur launch AI innovation lab

The new facility will link students and faculty to industry problems in healthcare, education technology and finance, as India pushes applied AI research.

By Sofiah Nichole Salivio • 4 min read • Last week
ThreatBook launches AI security tools for SOCs & AI skills
Security Operations Centres

ThreatBook launches AI security tools for SOCs & AI skills

Security teams could cut alert backlogs, while enterprises gain a way to inspect AI skills for hidden tampering and backdoors.

By Mark Tarre • 5 min read • Last week
JFrog unveils Mumbai speaker line-up on AI software risks
DevOps

JFrog unveils Mumbai speaker line-up on AI software risks

Indian firms are moving to tighten software controls as AI agents and code generation raise new security and auditability risks.

By Mark Tarre • 4 min read • Last week
KnowBe4 adds AI secure coding training with partner
Data Protection

KnowBe4 adds AI secure coding training with partner

Organisations using AI-assisted development can now get specialist secure coding training as KnowBe4 expands its library for technical teams.

By Joseph Gabriel Lagonsin • 4 min read • Last week