Our Scans · (ZY.4.04) Cognitive Science · Weekly Summary

• [New] Amid rising concern over AI-enabled threats, Germany's BSI has issued new guidance to help organizations defend large language models against evasion attacks such as jailbreaks, prompt injections, and other adversarial inputs, offering practical steps to harden LLM deployments. EU DisinfoLab
• [New] Juniper Networks deployed AI-driven behavioral analytics that auto-remediate threats. openpr
• [New] Emerging threats from AI-powered systems, such as the Chinese-developed Claude AI, are prompting advancements in cyber defence technologies. KeepTrack
• [New] 58% of CISOs expect an increase in vendor-origin cyber incidents in 2026, particularly involving SaaS, shared cloud platforms, APIs, and AI-driven services. Big News Network.com
• [New] Hospitals struggle to cut off compromised vendors and AI platforms, as the new 2026 US Hospital Cyber Readiness report from Black Book Research delivers actionable intelligence on kill-switch deficits, slow revocation timelines, and contract / insurance misalignment with third-party and AI risk. Big News Network.com
• [New] Imagine a world where startups can harness tools like OpenTelemetry and AI agents to build groundbreaking applications, all while minimizing risks and focusing on innovation. Coaio Limited
• [New] By 2030, generative AI is being touted to be deeply integrated into all levels of business, providing automation for everything from content generation to customer service and product design. Themes ETFs
• [New] Over 40% of U.S. work will be augmented or reinvented using automation and AI technologies. Isometrik.ai
• [New] The U.S. is in a compute arms race with China, and Washington's ability to access large, secure AI supercomputing clusters will influence everything from defence simulations to fraud detection and healthcare analytics. Tech Startups - Tech News, Tech Trends & Startup Fundin
• [New] Anthropic is scaling its Claude AI model on Microsoft Azure, powered by NVIDIA, which will broaden access to Claude and provide Azure enterprise customers with expanded model choice and new capabilities. RTInsights
• [New] The UK Government has announced the creation of AI Growth Zones, the first of which will be in Culham, Oxfordshire, a site that benefits from a major powerline connection that was previously used for an experimental nuclear reactor (now being decommissioned). Travers Smith
• [New] Google's DeepMind has unveiled an AI tool designed to autonomously detect and mitigate threats, a move that underscores the growing role of AI in cybersecurity. OpenTools
• [New] Innovations like Google's DeepMind platform that autonomously identifies and mitigates threats in real-time offer a glimpse into the future of cybersecurity, where AI battles AI to secure networks and data. OpenTools
• [New] AI-driven engagement will become mandatory for hospitals and healthcare networks across the globe. The Times Of India
• [New] British novelists are raising concerns that AI could replace their work, with nearly half saying the technology could 'entirely replace' them. Digital Watch Observatory
• [New] The economic impact of agentic AI systems could reach $15.7 trillion by 2030, with the most significant gains coming from productivity improvements in knowledge work and complex decision-making. Futurist Speaker on AI Leadership, Future of Work, Futu
• [New] By 2026, over 80% of enterprises will have deployed generative AI-enabled applications in production environments, up from less than 5% in early 2023. Futurist Speaker on AI Leadership, Future of Work, Futu
• [New] AI pioneer: Dubai is implementing an AI-powered system for monitoring and proactively maintaining infrastructure, which could reduce the cost of regular maintenance by 25%. Passle
• [New] Knowledge process outsourcing is accelerating at a 7.89% CAGR to 2030 as enterprises transfer analytics, regulatory reporting, and domain-specific research to Canada, where cultural proximity and language proficiency reduce context-setting cycles. Mordor Intelligence
• For 2026, we expect continued momentum, especially as AI becomes more context-aware and further integrated across mission-critical verticals such as venues and smart cities. asmag.com
• By 2026, agentic AI - systems that act, decide and adapt autonomously - will move from pilots to the core of how organizations operate and serve customers. Inside Telecom
• The next five years (2025-2030) are expected to bring unprecedented advancements, with AI agents potentially managing entire business functions with minimal human oversight, as predicted by AI Consultant Insights. Mixflow.AI

Last updated: 28 November 2025