AI Safety stories
Indian banks are bracing for higher losses as 84% of leaders report rising fraud and growing concern over AI-driven scams.
Businesses could soon verify and charge AI agents in milliseconds at the network edge, as autonomous traffic becomes harder to trust or block.
Agentic AI, zero-day surge, sovereign cloud, and humanoid robots will define IT strategy in 2027, Info-Tech Research Group warns.
Developers using AI assistants may get a verified knowledge base to cut repeated errors, security flaws and duplicated debugging work.
Businesses need a single view of AI agents as their access and ownership can change in real time across cloud and internal systems.
Many harmless prompts will now be diverted to Claude Opus 4.8 as Anthropic tightens safeguards around its newest general-use model.
AI-written database changes can now be checked and traced before deployment, as Liquibase Secure 5.2 targets production risk and audit gaps.
Most enterprises are still failing to turn agentic AI trials into usable gains, as weak governance and orchestration keep deployments in pilot mode.
The rollout aims to help customers tame rising AI-driven complexity as Datadog adds autonomous monitoring, security and agent oversight tools.
More than 1,000 CI&T AI engineers are being trained on Claude as the firm targets financial services, retail and consumer goods projects.
Customers facing machine-speed cyber breaches will get a single agent across Rubrik's platform, with auditable controls for autonomous recovery actions.
The funding will help the London- and New York-based software group expand its AI agents for revenue teams and hire globally.
Banks seeking compliant AI could gain tools that are easier to govern and audit as Titan uses fresh funding to expand its platform.
The wider rollout will put AI tools in the hands of more than 276,000 KPMG staff, as the firm pushes clients beyond pilot projects.
The rollout deepens Apple's AI push, but some users in Europe and China will miss key features as Siri's remake arrives later in beta.
Australia will get wider support to defend critical digital systems as Canberra and Microsoft deepen cooperation on cyber security and AI.
Boards are being warned to assess AI risk as well as opportunity, after new demand from executives prompted the course expansion.
The capital's lead in AI use may widen Britain's productivity divide, with many regional firms lacking the data and cloud basics to scale.
The telecoms group will use Anthropic's Claude Mythos Preview to spot vulnerabilities faster as cyber threats grow more automated.
Accessibility is moving into mainstream platforms, but hallucinations and privacy risks could still undermine users who rely on AI support.