OpenAI stories
Marketers can now plug AI agents into customer data and actions without Klaviyo's interface, as the firm widens access across its platform.
Logistics customers will be able to query stock and delivery data in plain English as Diamond tests a new AI agent for Despatchlab.
The update aims to cut AI costs and governance risks by reusing approved workflows and data inside tools such as ChatGPT and Copilot.
Growing alarm over AI safety has intensified after another former Google DeepMind engineer warned the technology could escape human control.
Enterprises under pressure to curb AI risk and cost are being given a hybrid option that blends generative models with OCR and rules-based extraction.
Downstream AI alerts in Australia and New Zealand are rising as agents and MCP links create fresh routes for sensitive data leakage.
Businesses can now run OpenAI's latest frontier model on AWS without managing infrastructure, as Bedrock gains access to GPT-6 Astra.
Borrowers could see more relevant credit offers as the platform uses live data and bank account insights to sharpen approvals.
Eligible financial firms can now tap their own DealCloud records in ChatGPT while keeping confidential data and access barriers intact.
Financial institutions will gain access to licensed market data and governance tools as OpenAI targets research, modelling and client documents.
Safety fears are deepening as more AI staff warn regulators may not keep pace with systems they say could soon outrun human control.
Chinese AI groups gain formal sway over PyTorch's direction, as developers seek support across more chips and cloud environments.
Businesses are now under pressure to turn early AI enthusiasm into measurable productivity gains, with governance and workflow redesign key to success.
Security teams under pressure from faster intrusions can now use Zscaler's new platform to spot and contain AI-driven attacks at machine speed.
Security teams are under pressure from faster AI-driven attacks, prompting Zscaler to add automated remediation and third-party integrations globally.
Security teams could cut manual triage as the new service ranks code and intrusion risks by real-world exploitability, not alert volume.
Calls for tighter AI safeguards have intensified after warnings the technology could outpace human control and threaten jobs, security and markets.
Fears are mounting that unchecked AI could unleash cyberattacks, market shocks and job losses long before any extinction event.
New Zealand users risk fake login pages and scam sites after ChatGPT search results were found pointing to unsafe links and downloads.
The deal will give more than 70,000 students and staff institution-backed AI access, alongside safeguards and AUD $1 million for sustainability research.
