AI Safety stories - Page 5

Businesses warm to AI agents haggling over deals, but consumers remain wary of handing software full control of spending.

Anthropic finds emotion-linked internal patterns in Claude Sonnet 4.5 can nudge blackmail and cheating, offering new clues for AI safety.

Niobium expands beyond hardware with encrypted cloud platform for private AI, aiming to let organisations process sensitive data without exposing it.

AI agents prompt accounting firms to redesign bookkeeping and tax workflows as junior staff tasks are automated and human oversight stays central.

Orange Business to weave Reality Defender's deepfake checks into enterprise communications for 7,000 customers amid rising fraud fears.

Kyndryl unveils a governance toolkit to help firms supervise autonomous AI agents across IT operations, as enterprises move beyond pilot projects.

F5 and Forcepoint have teamed up to link data discovery with runtime controls, aiming to curb AI risks as enterprises move systems into production.

KPMG survey finds most executives will keep pouring money into artificial intelligence through a recession, with USD $186 million planned on average next year.

ExpressVPN expands beyond its VPN roots with ExpressAI, a privacy-focused platform using confidential computing to keep prompts and chats hidden.

ExpressVPN expands beyond VPNs with encrypted AI chats, launching ExpressAI on confidential computing enclaves after an audit by cybersecurity firm Cure53.

Codenotary unveils AgentMon to help Chief Information Officers and security teams track AI agent behaviour, costs and policy risks.

Zapier adds AI Guardrails to its automation platform, giving teams inline checks to block risky content before it reaches business systems.

Animoca Brands' Yat Siu argues autonomous AI agents will become the internet's main actors, with blockchain supplying payments, identity and trust.

Gartner says explainable AI will help drive wider monitoring of generative AI systems, as firms seek stronger checks on bias, hallucinations and reliability.

Battleground-state voters want tougher health insurance rules and strong AI guardrails, with overwhelming support for mental health protections.

Boards lag on AI oversight as Diligent survey finds most directors use the tools, but few have embedded governance or risk controls.

New Relic has named Wendi Sturgis to its board as the observability software maker seeks sharper AI and international strategy oversight.

F5 and Forcepoint team up to give enterprises continuous AI security, linking data discovery with runtime controls to reduce risk in production systems.

OpenAI opens public Safety Bug Bounty to find agentic prompt injection, data exfiltration and other AI misuse risks.

F5 and Forcepoint have teamed up to link data governance with live runtime controls, closing security gaps across enterprise AI lifecycles.