Global Scans · Artificial intelligence · Weekly Summary

• [New] OpenAI, Anthropic, Google and Microsoft are among the AI builders warning about increasingly capable AI attacks, while security heavyweights including Cloudflare, CrowdStrike, Fortinet, and Palo Alto Networks have also signed on. theregister
• [New] In 2026, AI infrastructure spending is projected to reach $450 billion, with inference workloads accounting for over 70% of AI compute demand - marking a structural shift from training-dominated to inference-driven AI. ODG Electronics
• [New] The RIMC meeting in 2026 will focus on agentic AI systems which do not simply act as passive chatbots waiting for inputs but instead function as active agents capable of making autonomous decisions in order to solve complex tasks. News Powered by Cision
• [New] With AI adoption accelerating in 2026, its tokenomics for decentralized AI services could see demand surge. BMIC
• [New] Organizations are increasingly considering external support for AI-enabled customer-service capabilities, with executives expecting to outsource 85% of customer-support inquiries and 81% of customer communications. Persistence Market Research
• [New] Ongoing efforts aim to embed generative AI in future calls across Horizon Europe, DEP, and the EIC for 2026 and 2027, reinforcing Europe's position as a global hub for secure, inclusive, and innovative AI technologies. Shaping Europe's digital future
• [New] In 2026, AI chatbots will become more intelligent, responsive, and deeply embedded across educational platforms. Interface
• [New] Asia-Pacific is at an important stage in how it regulates technology in the age of AI, and the decisions made over the next few years will shape the safety of the ecosystem. Fintech Singapore
• [New] AI-powered data centers are becoming significant consumers of electricity and water, with demand expected to increase substantially by 2030. GCC BUSINESS WATCH
• [New] With worldwide spending on generative AI projected to surge significantly, the demand for skilled professionals across healthcare, marketing, and software sectors is growing rapidly. Research.com
• [New] AI will actively generate hypotheses and control scientific experiments in physics, chemistry, and biology. BuildEZ
• [New] 2026 will be an era of true collaboration between technology and people, with AI amplifying human expertise rather than replacing it. BuildEZ
• In the 2026 Thales Data Threat Report, 70% of organizations rank AI as top data security risk and 48% have experienced reputational damage as a result of AI-generated misinformation. Thales Cloud Security Products
• Insights reveal a resurgence in AI budget growth for 2027, following a more cautious approach in the previous year, signalling that firms are preparing to scale deployments across proven, high-value AI use cases. Default
• Memory demand is outpacing supply, with HBM and DRAM prices likely to remain elevated through at least 2027 as AI infrastructure expands. Seeking Alpha
• AI is poised to shrink the world's health gap and accelerate discovery in physics, chemistry, and biology, leading to more personalized learning experiences and a future where every research scientist could have an AI lab assistant. TashiOS
• 86% of companies anticipate an increase in their AI budgets in 2026. TashiOS
• By 2030, AI agents are projected to be responsible for 15-25% of all US e-commerce purchases. J.P. MORGAN
• There is now an increased risk of AI tools spiraling out of control. BBC News
• The World Economic Forum's Global Cybersecurity Outlook 2026 surveyed more than 1,200 cyber leaders and found that AI-driven attacks now sit at the top of the CISO threat priority list for the first time. Swif.ai
• The first half of 2026 marked the first time TrendAI research teams documented China-aligned threat actors Earth Krahang, Earth Naga (associated with Flax Typhoon), and an unidentified actor actively integrating generative AI into multiple stages of the attack lifecycle. TrendAI

Last updated: 27 September 2026