360 Launches nanoWork Enterprise Edition Channel Ecosystem with Native Security for AI Agents
On August 17, the 360 nanoWork Enterprise Edition partner recruitment launch event successfully convened, bringing together more than 200 channel partners from across China in a virtual gathering. This marked the first major assembly of channel partners since the product’s official release on July 28.
360 Group founder Zhou Hongyi stated at the event: “AI is moving from ‘being able to answer questions’ to ‘being able to complete tasks.’” The recruitment launch represents the critical step in turning this vision into reality by ensuring AI can truly “do work and deliver results,” supported by a security foundation, channel network, and service system.
360 Commercial Comprehensive Channel Department General Manager Wang Yi noted that the ToB enterprise digitalization sector is undergoing a definitive iteration, with AI agents replacing traditional office software as the next-generation productivity tool. The industry has shifted from conceptual observation to commercial deployment. Despite the trillion-dollar blue ocean opening, enterprise AI implementation remains challenging, with AI agent penetration in business operations at only about 18% and over 70% of AI projects failing to reach production environments.
Security as the Foundation
Wang Yi explained that 360 leverages more than 20 years of security accumulation to position nanoWork Enterprise Edition as a truly deployable enterprise-grade AI agent platform that improves efficiency and is easy for channels to sell and deliver. The product features “native security, built-in from the factory,” embedding 360’s two-decade security expertise and national-level cyber defense capabilities through mechanisms such as cloud isolation, permission control, and data protection. This allows enterprises to confidently assign real work to AI by controlling what the AI can see, do, and share results with.
Channel Architecture and Revenue Models
The company released its latest channel policy featuring a three-dimensional partner architecture, four profit models, and a five-tier empowerment system. The three-dimensional framework uses “classification × layering × grading”: classification divides partners into regional, industry, ecosystem strategic, and scenario co-construction categories; layering distinguishes between agency and joint-operation models; grading certifies core, key, and potential tiers with increasing benefits. Four core revenue paths cover account expansion, compute consumption, training services, and陪跑 delivery. The five empowerment systems provide training, certification, content, tools, and sales support through online live sessions, on-site training, certification bootcamps, dedicated learning platforms, and real-time sales assistance groups.
Service and FDE Network
Service remains central, supported by a 1,000+ frontline deployment engineer (FDE) legion and the AikerWorld community. FDEs, drawn from 360’s “white hat” teams, deliver six service modules including AI consultation, product training, customization, deep development, cloud陪跑, and on-site驻场 support. Two delivery modes are offered: an陪跑 mode bundled with product seats for guided adoption and standalone service packages for deeper implementation. The AikerWorld community connects AI developers, industry experts, and partners with PMI China’s 680,000 certified project managers to scale the expert network nationwide.
Through city-level channel partners, 360 combines headquarters capabilities with localized service to bring enterprise AI into real business scenarios across enterprises, stores, and industrial parks.
Related articles
OpenAI GPT-6 Astra Deploys Multi-Agent Parallel Processing, Increasing Local CPU Load and Security Risks
Early users of GPT-6 Astra have observed the model distributing complex tasks across multiple specialized agents that plan, solve, test code, verify results, and iterate after failures. This multi-agent approach enables faster handling of multi-step workflows compared to sequential chatbots. OpenAI states that Astra can control computers, operate browsers and applications, and install or test software, though it has not officially confirmed a native multi-agent architecture. Main computations run in the cloud, but agent tools can execute on user devices or corporate servers, leading to noticeable processor load when multiple agents compile code, launch browsers, run tests, and operate containers simultaneously. Corporate environments face added complexity as each agent requires virtual machines, sandboxes, internal data access, and careful environment cleanup. The increased autonomy has prompted OpenAI to strengthen monitoring of Astra actions and permission boundaries for subscribers of ChatGPT and enterprise clients.
Microsoft Copilot Can Surface Overshared Data Despite Permission Boundaries
Microsoft documentation states that Copilot only accesses data authorized for the signed-in user, yet default SharePoint and OneDrive sharing settings often grant broad access that the AI then respects literally. This creates accidental oversharing risks where Copilot retrieves documents shared too widely years earlier. Administrators can use Content Management Assessment and Data access governance reports, including the EEEU report covering the top 100 sites shared in the past 28 days, to identify problematic content. Two distinct controls exist: Restricted Access Control removes access entirely while Restricted Content Discovery hides items from Copilot and search without altering permissions. Sensitivity labels combined with encryption can exclude programmatic access for agents, though Microsoft does not guarantee outright blocking. Interaction logs stored in Microsoft Purview retain user prompts, Copilot responses, and citations to accessed documents, providing an audit trail for oversharing incidents.
Adaptive LLM Worm Uses Local Models to Craft Per-Target Exploits in Heterogeneous Networks
Researchers from the University of Toronto have published a preprint describing an adaptive computer worm driven by LLM agents that spreads across corporate networks by generating individualized attack strategies for each compromised system. Unlike traditional worms such as WannaCry that rely on fixed exploits, this worm maintains its own infrastructure by running local LLMs on infected GPU-equipped machines to analyze vulnerabilities and synthesize new attack vectors in real time. The system was tested in an isolated FakeCorp environment containing Linux, Windows, and IoT devices, successfully leveraging known real-world vulnerabilities to propagate over 48 hours and seven-day autonomous runs. Two core components power the worm: a GPU-hosted LLM component and a hierarchical agent framework with memory, reasoning graph, and tool modules that manage reconnaissance, exploitation, and payload deployment. The authors note that the approach creates an economic asymmetry favoring attackers because the worm parasitizes victim compute resources, eliminating the need for external C2 or commercial LLM services. They warn that adding adaptive reasoning to historical worms such as SQL Slammer, Conficker, or Stuxnet would significantly increase their resilience while remaining slower and noisier than classic self-propagating malware.
Building Secure On-Prem AI Assistants: How to Keep Corporate Data Inside Closed Contours
Many organizations hesitate to deploy AI assistants due to strict data protection rules that prohibit sending information to external clouds. The article explains how to implement AI models entirely within a company's own infrastructure, ranging from on-premise servers to fully offline laptops. It breaks down four deployment locations from public APIs to local devices and clarifies three distinct access levels: read, write, and execute. The author emphasizes that most business value comes from read-only access combined with human-in-the-loop controls for any irreversible actions. Practical recommendations include RAG over model size, quantization for local hardware, and maintaining immutable audit logs. The piece also warns that preparing clean knowledge bases often consumes more effort than the model itself.