Introduction
The second full week of February 2026 brought clear momentum in enterprise AI platforms, open-source modelling breakthroughs, and wider industry competition. From platforms making AI agents first-class workers to new high-performance open models emerging from China’s AI labs, this week showed how global innovation and adoption continue to accelerate.
1. OpenAI Launches Frontier to Operationalise AI Agents in Enterprise
OpenAI unveiled Frontier, a major new enterprise platform designed to help organisations build, deploy, and manage AI agents that operate like digital coworkers — from shared onboarding and permissions to integrated workflow execution. The platform aims to address a key challenge in enterprise AI: turning isolated pilots into scaled, governed productivity tools across departments and business systems. Early adopters include HP, Oracle, State Farm and Uber.
2. z.ai GLM-5 Sets New Benchmark for Open-Source Models
Chinese AI powerhouse Zhipu AI — internationally known as z.ai — released GLM-5, a new open-weight large language model with ~744 billion parameters, achieving a record-low hallucination rate and competitive reasoning and coding performance compared with major proprietary models. The model is designed to power agentic workflows, reasoning tasks, and open commercial use under an MIT licence, challenging incumbent AI leaders and deepening global open-source innovation.
3. Enterprise Confidence in AI Infrastructure Remains Strong
Industry analysts highlighted robust AI capital expenditure plans for 2026 as hyperscalers and enterprise vendors forecast continued growth in AI infrastructure investment, with aggregate plans approaching $700 billion — reflecting demand for data-centre capacity, custom accelerators and agent platforms that support next-generation AI workloads.
4. AI Helps People Navigate Essential Services — New Research
A study involving researchers from The University of Manchester found that AI systems could significantly help individuals who feel overwhelmed or excluded by essential administrative tasks — such as paying bills or accessing healthcare services — by acting as personal navigators for everyday life challenges. This research underscores how AI continues to expand into social and accessibility use cases beyond traditional enterprise and consumer applications.
5. ETSI AI & Data Conference 2026 Highlights EU Standardisation Work
The ETSI AI and Data Conference 2026 took place in France, focusing on how AI and data technologies can be brought together to support European AI policy objectives and standardisation under the EU’s broader regulatory frameworks. Discussions stressed interoperability, governance mechanisms and compliance — showing that EU AI governance continues to shape how international tech development is aligned with policy.
Conclusion
This week’s AI news underscores a transition from experimentation to industrialisation: platforms like OpenAI’s Frontier are making agents practical business assets, open-source models like GLM-5 are driving performance competition, and both research and policy communities are shaping how AI is embedded in society and enterprise. As 2026 progresses, expect these threads — platformisation, open collaboration, and governance — to remain central.
