AI Safety stories - Page 23

AI agents with sweeping powers will trigger new identity security fears by 2026, as firms brace for over-privileged bots and poisoned models.

AI rollout delays are stretching to a year as three-quarters of organisations report security incidents and governance gaps grow.

DXC launches AdvisoryX to help enterprises scale AI, as 94% struggle to move beyond pilots and build coherent business cases.

AI data leaks are helping drive a global cyber attack surge, with firms now hit by over 2,000 assaults a week and ransomware on the rise.

Arctic Wolf predicts agentic AI will overhaul SOCs, tighten Zero Trust and keep humans central as cyber risk surges into 2026.

AI-powered attackers and defenders will clash in 2026, driving autonomous breaches, VPN failures and stricter rules that reshape cyber security.

Netskope adds Model Context Protocol controls to its One platform, giving security teams new tools to monitor and lock down AI agent traffic.

Check Point updates Quantum firewall with AI-aware controls, aiming to curb data leakage, model abuse and new automated cyber attacks.

Enterprises must ditch open-world hype and build deterministic, event-driven AI agents that reliably automate closed-world workflows at scale.

AI supply chains will consolidate by 2026, with tight human oversight, integrated infrastructure and trust becoming key differentiators.

Amazon launches Nova 2 AI suite with four new models and introduces Nova Forge for firms to create custom AI solutions tailored to diverse industries and tasks.

EY has launched a physical AI platform with NVIDIA tech and opened an EY.ai Lab to advance AI in robotics, drones, and smart devices across industries.

Norton has launched Neo, a free AI-native browser offering enhanced privacy and security against rising AI-driven online threats, now available worldwide.

Anthropic launches Claude Opus 4.5, enhancing coding, reasoning, and security with improved performance and resistance to cyber attacks.

AI red teaming blends security tactics with safety goals to prevent exploits in chatbots, defending users from harm beyond classic cyber threats.

Gluware unveils Titan, an AI validation platform ensuring assured, compliant, and reversible automation for complex enterprise networks.

OpenAI's AI models lead secure code generation with up to 72% pass rate, outpacing rivals who show little progress despite ongoing sector development.

Invisible AI failures, such as hallucinations and accuracy issues, threaten enterprise trust, with 82% of bugs traced to these hidden errors, Testlio finds.

Google launches Gemini 3 AI with advanced multimodal and reasoning capabilities, enhancing tasks from research to complex coding across multiple platforms.

AI firms like Copilot and OpenAI set new efficiency records, generating millions in revenue per employee, highlighting the sector's rapid growth and lean workforces.