The global AI agent market is no longer a singular, borderless frontier. This week, three distinct developments across Asia-involving Manus, Enhans, and Huawei-demonstrate that the next phase of agentic deployment is being shaped by regional capital, sovereign infrastructure, and heavy-handed regulatory oversight. While US-based firms like Sierra and Cognition continue to command massive valuations, the landscape is fracturing into regional silos that prioritize domestic control over global interoperability.
The most striking example of this shift is the trajectory of Manus. According to Bloomberg, the company is in advanced discussions for a $500 million funding round at a $4 billion valuation. This follows a turbulent period where Beijing’s NDRC blocked Meta’s $2 billion acquisition of the firm in April 2026. The regulatory intervention was absolute: co-founders Xiao Hong and Ji Yichao were summoned and barred from leaving China, effectively forcing the company to pivot from a potential US acquisition target to a cornerstone of the domestic Chinese AI ecosystem. While this valuation represents a significant recovery, it is important to note that the round is not yet closed, and terms remain subject to change.
In Seoul, the narrative is defined by deep-rooted industrial integration rather than regulatory friction. Enhans, which recently raised a $38 million Series C (confirmed via a September 17, 2026 press release), is moving beyond the typical startup growth model. The round was co-led by TIMEFOLIO Asset Management and Stonebridge Ventures, but the strategic participation of POSCO Investment, LG CNS, and Lotte Ventures is the real story. These chaebol groups are not just investors; they are already running Enhans’ technology in live production. By utilizing ontology-based data structuring to bridge disparate SAP, ERP, and CRM systems, Enhans is embedding its AgentOS directly into the backbone of Korean manufacturing and finance. Their ACT-2 agent, which leverages screen-level automation to operate software and websites the way a human user would, highlights how regional players are solving the “last mile” of enterprise automation.
Meanwhile, Huawei is building the sovereign infrastructure required to host these agents at scale. At HUAWEI CONNECT 2026, the company launched its AI Cluster Service (AICS) and a suite of agent-specific tools, including Agentic MaaS and the AgentArts platform. Dr. Peter Zhou’s keynote emphasized that infrastructure must now evolve to support secure, reliable agent workloads in production. Huawei’s approach is comprehensive, spanning from silicon to a platform that already serves over 3,500 customers through its Agentic Infra paradigm. While AICS is available in China starting September 30, global availability is not slated until November 30, marking a clear timeline for the expansion of their sovereign cloud stack.
These moves collectively signal that the agent market is splitting. We are seeing a divergence in how data is structured, where compute is hosted, and which regulatory frameworks govern autonomous actions. Enterprise buyers are no longer choosing between generic models; they are choosing between regional ecosystems. A company operating in Seoul faces different integration requirements than one in Shanghai or Silicon Valley, particularly as infrastructure providers like Huawei bake specific memory and scheduling paradigms into their cloud offerings.
This fragmentation echoes the challenges we have previously tracked in our coverage of The Shadow Agent Gap and the fragmented landscape of early-stage AI, such as our reporting on Twin1 AI and Salesforce in Claude. Just as the push for enterprise data portability highlighted the friction between closed platforms, these regional developments suggest that “agentic interoperability” may become the primary technical hurdle of the next decade. If an agent built on a Huawei-optimized stack cannot easily communicate with an agent running on a Western cloud, the promise of a seamless, global agent economy will remain elusive.
The open question is whether these regional silos will eventually converge through open standards or harden into permanent, incompatible zones. For now, the capital flows and infrastructure deployments suggest that the latter is more likely. As enterprise buyers navigate this landscape, the decision of which agent ecosystem to adopt will be as much about geopolitical alignment as it is about technical capability.