AI Safety stories
The chipmaker's technology is now embedded across robotaxi development, underscoring its role in a market Nvidia says could hit USD $400 billion by 2035.
Enterprises risk exposing sensitive records as AI agents inherit broad access by default, leaving accountability and audit trails unresolved.
Widespread use of chatbots and drafting tools may leave workers less able to verify outputs or think independently, Walsh said.
Banks are broadening use of RDC.AI's software as the Australian-founded group reports more than 120 per cent growth in existing customers.
Developers can now access cheaper GPT-6 options as OpenAI cuts API prices by half and says caching changes will speed up agent workflows.
Lower running costs and faster output are aimed at businesses using Anthropic's latest Claude model, which starts at USD $4 per million input tokens.
The breach highlights how AI agents can keep probing blocked systems, prompting calls for stronger defences around identities, networks and data.
Borrowers could get funds on a Friday night as LoanPro folds real-time card disbursement, payments and AI servicing into one platform.
Spanish-speaking voters could be left with weaker guidance after researchers found leading AI chatbots gave flawed election answers across 10 US states.
Product teams are being pushed to scrutinise AI outputs and governance as Scrum.org adds a five-hour self-paced course with certification.
The badge strengthens customer confidence as firms racing to deploy Copilot and other AI tools face tighter scrutiny over data access and governance.
AI tools are exposing smaller firms to faster scams and data leaks, making basic controls and staff training vital to stay resilient.
Singapore's scarce land and tight labour market are pushing H2O.ai to test AI systems in live customer settings, not demos.
Public pressure for AI guardrails is surging faster than it did for social media, as Australia joins a growing coalition urging action.
Paid Jira customers will get tighter controls and better measurement of AI agents as Atlassian widens software teams' access to shared project context.
Weak controls left Australia's Medicare database exposed after an AI agent was able to press past access restrictions, experts say.
Australia's AI rules may decide whether businesses gain protection without extra red tape that slows deployment and pushes investment offshore.
Blind spots in agentic AI could turn routine tasks into breaches, as Canadian firms ramp up use faster than oversight can keep pace.
Despite strong interest in AI shopping tools, most Singapore consumers want to approve each purchase themselves, a Global Payments survey found.
The reported incident has heightened fears that autonomous AI agents could probe critical systems faster than governments can contain them.
