AITECH Cloud Network is putting privacy at the center of AI agent development.
The company, formerly known as Solidus AI Tech, has partnered with Secret Network to integrate confidential computing into Agent Forge, its no-code platform for building and deploying AI agents and automated workflows.
The collaboration is aimed at one of enterprise AI’s biggest unresolved problems: how to give businesses access to increasingly capable AI models without forcing sensitive data into environments they cannot fully control.
Under the partnership, Agent Forge users will gain access to confidential AI models and privacy-focused workflow execution powered by Secret Network’s confidential computing infrastructure. The companies also plan to add verification tools so users can confirm that AI workloads are actually running inside trusted confidential environments.
For enterprises that want AI but have spent years building security policies around sensitive data, that’s potentially more important than another chatbot feature.
Agent Forge gets a privacy upgrade
Agent Forge launched in April 2025 after being developed internally by AITECH Cloud Network.
The platform is designed to let users create AI agents, workflow automations and integrations without writing traditional code. It offers templates and connections to external services, with the goal of making agent development accessible to users ranging from AI newcomers to enterprise teams.
The Secret Network integration adds a different layer to that proposition.
Instead of focusing solely on what an AI agent can do, the partnership addresses where and how the agent processes data.
Initial integration work will focus on Agent Forge’s light-mode chat interface. The companies plan to extend confidential computing capabilities into the broader workflow builder as development progresses.
That could eventually allow users to construct AI workflows in which sensitive information remains protected while the AI performs its assigned tasks.
The distinction is important because AI agents increasingly need access to business data to be useful.
An agent that can only operate on public information is relatively easy to secure. An agent that can access customer records, proprietary documents, financial information or internal company systems is far more valuable—and far more difficult to trust.
What confidential computing brings to AI
Confidential computing is designed to protect data while it is being processed, rather than only when it is stored or transmitted.
That matters because traditional security models tend to focus on data at rest and data in transit. Once information is being processed by an application, it must generally be accessible to the underlying computing environment.
Confidential computing introduces hardware- or cryptography-backed mechanisms designed to isolate sensitive workloads and restrict unauthorized access during processing.
In AI, that can be particularly useful.
AI models increasingly need access to proprietary information. Businesses may want to analyze internal documents, customer data or operational information without exposing that material unnecessarily to infrastructure providers, third parties or other workloads.
The challenge is to make the AI useful without turning sensitive data into an uncontrolled input.
Secret Network’s technology is intended to address that problem by providing confidential computing infrastructure for applications that need privacy-preserving execution.
For Agent Forge, the result could be a platform where users can build AI agents that operate on sensitive information while maintaining stronger controls around how that information is processed.
Verification is the more interesting piece
Privacy alone isn’t enough for enterprise AI.
Companies also need to know whether the privacy claims can be verified.
That’s why the partnership’s planned verification tools could be particularly significant.
AITECH Cloud Network and Secret Network say users will be able to confirm that AI workloads are running within trusted confidential computing environments.
That introduces an important concept: verifiable execution.
In a conventional cloud environment, customers largely trust the provider’s infrastructure and security controls. Confidential computing can strengthen those assurances by providing technical mechanisms that help prove where and how workloads are running.
For AI agents, that becomes increasingly relevant as they gain the ability to execute multi-step workflows autonomously.
An enterprise may be willing to let an AI agent summarize a public document without much concern. Giving that same agent access to confidential contracts, customer information or proprietary research is another matter entirely.
Verification can help establish a stronger trust boundary.
It doesn’t magically make an AI system secure, but it can give organizations additional evidence that their workloads are being processed within the environment they expect.
Why enterprise AI needs more than model performance
The AI industry has spent the past several years obsessing over model performance.
Which model is smarter? Which produces better reasoning? Which handles longer contexts? Which agent completes tasks more reliably?
For enterprise buyers, those questions matter.
But they are only half of the equation.
An AI model can be incredibly capable and still be unsuitable for a particular company if the organization cannot safely provide it with the data it needs.
This is one reason privacy-preserving AI infrastructure is gaining attention.
The next phase of enterprise AI is likely to involve more sensitive workloads. AI agents won’t just generate marketing copy or answer employee questions. They will increasingly interact with internal databases, financial systems, customer records and proprietary intellectual property.
That creates a paradox.
The more useful AI becomes, the more sensitive the information it needs.
Confidential computing is one potential answer to that problem.
Agent Forge is taking the no-code approach to agents
The partnership also highlights where Agent Forge is trying to differentiate itself.
There is no shortage of AI agent frameworks. Developers can build agents using model APIs, orchestration frameworks, workflow tools and cloud services.
Agent Forge is taking a more accessible route.
Its no-code environment is designed to let users build agents and workflows without traditional software development. That lowers the barrier for businesses that want to experiment with agents without building an entire AI engineering team.
Adding confidential computing to that environment could make the proposition more attractive to organizations that have been hesitant to experiment with AI because of data-security concerns.
The important word is could.
Enterprise adoption will ultimately depend on how the privacy technology performs in real workloads, how easy the controls are to use, what models and integrations are supported, and whether organizations can satisfy their own compliance requirements.
Security that requires a Ph.D. and a week of configuration isn’t exactly frictionless.
Agent Forge’s challenge will be to make confidential AI feel as simple as the rest of its no-code experience.
Ethereum migration adds another layer
Agent Forge has also completed its migration to Ethereum, according to AITECH Cloud Network.
That move is notable because the company is positioning the platform for broader institutional adoption.
The combination of AI agents, blockchain infrastructure and confidential computing places Agent Forge at the intersection of several technology trends that are increasingly converging.
Blockchain can provide coordination, identity, payments or verification mechanisms.
AI agents provide automation and decision-making.
Confidential computing protects sensitive processing.
None of those technologies automatically makes an enterprise application better. But together, they can address some of the structural problems that emerge when autonomous software begins interacting with real-world systems and valuable data.
The industry is still figuring out the architecture.
Agent Forge’s strategy suggests AITECH Cloud Network wants to participate in that emerging stack rather than compete solely on AI agent functionality.
Secret Network is betting on confidential AI
For Secret Network, the partnership expands the potential reach of its confidential computing technology into a no-code AI environment.
That matters because confidential computing has historically been associated with infrastructure, security and specialized development environments.
Bringing those capabilities into an AI agent platform makes them more accessible to businesses that may not want to become experts in privacy-preserving infrastructure.
The partnership also reflects a broader shift in how AI security is being discussed.
Early enterprise AI conversations focused heavily on preventing users from accidentally leaking data into public models.
The newer challenge is more complex.
Organizations want AI systems that can actively work with sensitive information while maintaining control over that information throughout the workflow.
That’s a fundamentally different requirement.
It’s the difference between telling employees, “Don’t put confidential documents into the chatbot,” and building a system where confidential documents can safely be used by an AI agent under defined controls.
The second approach is much more powerful.
It’s also much harder.
The rise of confidential AI
The partnership arrives as the AI industry increasingly explores confidential computing and trusted execution environments as part of enterprise AI infrastructure.
Major cloud providers have been developing confidential computing capabilities for years, while chipmakers have introduced hardware designed to isolate sensitive workloads.
The concept is becoming increasingly relevant to AI because models are moving closer to proprietary business data.
There is also a regulatory dimension.
As governments increase scrutiny of AI systems, organizations are under growing pressure to understand where data goes, who can access it and how automated systems operate.
Privacy-preserving infrastructure does not solve every regulatory problem, but it can become part of a broader compliance and governance strategy.
That makes confidential AI less of a niche security feature and more of an architectural consideration.
The hard part: proving the entire workflow is private
There is, however, an important distinction between confidential model execution and a fully confidential AI workflow.
An AI agent rarely operates in isolation.
It may retrieve information from an external database, call another API, interact with a business application, send a response to a user and trigger another agent.
Protecting one component doesn’t necessarily protect the entire chain.
That means the planned expansion from confidential model access into fully private AI workflows will be the more meaningful phase of this partnership.
The companies say future deployments will include confidential execution environments capable of supporting fully private AI workflows.
If that can be delivered without forcing users to redesign their applications around complex security infrastructure, Agent Forge could have a compelling proposition.
If confidentiality stops at the model boundary, its usefulness will be narrower.
The architecture will matter.
AI agents make privacy harder—and more important
The timing of the partnership is no accident.
AI agents are moving from experimental assistants toward software that can take actions.
A conventional chatbot waits for a question and generates a response.
An agent can potentially retrieve information, make decisions, call tools and perform tasks across multiple systems.
That autonomy changes the security equation.
An agent with access to business systems can potentially expose more information, make more consequential decisions and interact with more external services than a basic chatbot.
Security therefore has to be designed around the entire agent lifecycle.
Who gave the agent permission?
What data can it access?
Where is that data processed?
What actions can it take?
Can those actions be audited?
Can the organization verify the environment in which the workload ran?
Those are the questions enterprise customers are likely to ask as agents move into production.
Agent Forge’s partnership with Secret Network is an attempt to answer at least some of them at the infrastructure level.
A privacy layer could become a competitive advantage
For no-code AI platforms, ease of use has traditionally been the headline feature.
The pitch is simple: build an AI agent without needing a team of engineers.
But as businesses move from experimentation to production, security and governance can become equally important.
A platform that lets a marketing team build an agent in an afternoon is useful.
A platform that lets the same team build an agent that can safely process sensitive customer data under enterprise security controls is potentially much more valuable.
That is the opening AITECH Cloud Network appears to be pursuing.
Rather than making confidentiality a separate infrastructure project, it wants to make it part of the Agent Forge experience.
If successful, that could change the conversation from “Can we build this agent?” to “Can we safely deploy this agent?”
The latter is the question that determines whether many AI experiments ever reach production.
What comes next
The initial integration will focus on confidential model access and verification features.
After that, the companies plan to introduce confidential execution environments for private AI workflows.
They also expect to collaborate on community initiatives, educational content and ecosystem development.
For users, the most important milestones will be practical rather than promotional.
Can Agent Forge protect sensitive information without adding significant workflow complexity?
Can customers independently verify that workloads are executing within the promised confidential environment?
How many models and external services will support the confidential architecture?
And can confidential agents operate with enough speed and flexibility for production workloads?
Those answers will determine whether the partnership becomes a meaningful enterprise capability or simply another checkbox on the increasingly crowded AI platform feature list.
Privacy may become the new AI battleground
The AI industry’s first competitive race was about intelligence.
Then it became about cost and speed.
Now, as AI agents begin interacting with sensitive enterprise data, trust is becoming just as important.
AITECH Cloud Network and Secret Network are betting that confidential computing can provide part of that missing trust layer.
The strategy makes sense on paper.
Agent Forge brings the user-friendly agent-building environment. Secret Network brings confidential infrastructure and verification capabilities. Together, they are targeting businesses that want autonomous AI workflows without giving up control over sensitive information.
The real test will come when those workflows move beyond demonstrations and into production.
Enterprise AI adoption isn’t blocked solely by a lack of capable models. Increasingly, it is constrained by whether companies can deploy those models safely.
If confidential computing becomes a standard requirement for sensitive AI workloads, partnerships like this one could become increasingly important.
For Agent Forge, the goal is straightforward: make private AI as easy to build as public AI.
That’s a much harder engineering problem than adding another chatbot.
It may also be the problem that determines which AI agent platforms enterprises are willing to trust.
