As artificial intelligence accelerates across every part of the business, AI governance has shifted from a trend to a necessity. As AI systems become more integrated into our daily lives, from healthcare diagnostics to autonomous vehicles, ensuring they are developed and deployed responsibly is paramount. One compelling framework for tackling this challenge is the layered model for AI governance, which provides a structured way to address the multifaceted issues surrounding AI. In this blog post, we’ll explore this model, break it down into its core components, and connect it to real-world applications, particularly in the realm of trade secrets. We’ll also spotlight how platforms like Tangibly are making these abstract concepts more actionable for businesses.
Understanding the Layered Model for AI Governance
The layered model for AI governance, as proposed by researchers Urs Gasser and Virgílio A.F. Almeida, offers a conceptual framework to bridge the information gaps between AI developers, users, and policymakers. This model recognizes that AI isn’t governed in a vacuum; instead, it operates across interconnected layers that encompass social, ethical, legal, and technical dimensions. By thinking in layers, we can systematically tackle challenges like bias, transparency, and accountability without overwhelming any single aspect.
At its core, the model is divided into three primary layers:
1. Social and Ethical Layer:
This foundational layer deals with the broader societal implications of AI. It includes norms, values, and ethical guidelines that shape how AI should align with human rights and social good. For instance, questions like “How does AI impact privacy?” or “What biases might it perpetuate?” are addressed here. Organizations like the IEEE and OECD have contributed guidelines that fit into this layer, emphasizing fairness and inclusivity.
2. Legal and Regulatory Layer:
Building on the ethical base, this layer focuses on formal rules, laws, and policies. It encompasses regulations such as the EU AI Act or the U.S. National AI Initiative, which classify AI systems by risk levels and impose requirements for high-risk applications. This is where compliance comes into play, ensuring AI deployments adhere to data protection laws like GDPR or intellectual property statutes.
3. Technical Layer:
The innermost layer involves the nuts and bolts of AI systems: algorithms, data sets, and infrastructure. Here, governance translates to practical tools like auditing mechanisms, explainable AI techniques, and security protocols to make systems robust and verifiable.
This layered approach isn’t hierarchical in a strict sense; rather, it’s interdependent. Changes in one layer ripple through others, creating a holistic governance ecosystem. For example, ethical concerns (social layer) might drive new regulations (legal layer), which in turn require technical innovations (technical layer).
Applying the Layered Model to Trade Secrets: Five Operational Layers
To make the model applicable to trade secrets in AI, it extends into five focused layers that integrate policy, technology, and processes.
Layer One: Clear Policies for What Can and Cannot Enter AI Systems
Organizations need rules for employee inputs to AI systems, including:
-
- Prohibited information categories for generative AI.
- Approval pathways for sensitive materials in R&D, product, and legal teams.
- Team-specific guidance for engineering, research, operations, and business development.
These define confidentiality and trade secrets. Tangibly provides a unified inventory with ownership and access rules for clarity.
Layer Two: Technical Controls and Access Management
Technical safeguards enforce policies, such as:
-
- Restricting external AI in high-risk areas.
- Internal sandboxes to block sensitive inputs.
- Logging AI interactions.
- Granular access to assets.
Data access secures AI models. Tangibly ties controls and logs to the inventory for protection.
Layer Three: Continuous Identification and Classification of Trade Secrets
Identify undocumented secrets in patents, notes, code, contracts, and materials. This high-risk, high-value information needs mapping. Tangibly’s AI scans and extracts know-how, creating a real-time asset map to guide governance.
Layer Four: Governance Processes that Scale
Repeatable workflows include:
-
- Updating inventories.
- Reassessing AI tools.
- Approving partnerships.
- Onboarding and offboarding.
- Aligning teams.
This ensures dynamic governance. Tangibly offers workflows for ongoing classification and updates.
Layer Five: Incidents, Enforcement, and Legal Preparedness
Handle exposures with:
-
- Trade secret documentation.
- Protection evidence.
- Access chains.
- Investigation tools.
This supports legal actions. Tangibly supplies metadata and evidence from governance data.
Making AI Governance Tangible: The Role of Trade Secrets
While the layered model provides a high-level blueprint, applying it in practice can feel abstract. This is where we need to make governance tangible: grounding it in real tools and strategies that organizations can implement today. One critical area where this manifests is in the protection of trade secrets, which are increasingly vital in the AI era.
AI development often relies on proprietary algorithms, training data, and models that companies guard fiercely. Unlike patents, which require public disclosure, trade secrets allow firms to maintain confidentiality, providing a competitive edge in fast-paced fields like machine learning. However, managing these secrets poses governance challenges across all layers of the model:
-
- Social and Ethical Layer: Trade secrets must balance innovation with transparency. Overly secretive practices can erode public trust, especially if AI systems influence societal outcomes like hiring or lending decisions.
- Legal and Regulatory Layer: Laws like the Defend Trade Secrets Act (DTSA) in the U.S. provide frameworks for protection, but enforcement requires robust documentation and risk assessments. In AI, where data flows across borders, compliance with international regulations adds complexity.
- Technical Layer: This involves implementing access controls, encryption, and monitoring tools to prevent leaks or reverse engineering.
The layered model highlights how trade secret management isn’t isolated: it’s a thread weaving through ethical considerations, legal obligations, and technical safeguards. Poor governance here can lead to intellectual property theft, stifling innovation and raising ethical red flags.
Spotlight on Tangibly: Bridging Theory to Practice
To make the layered model more than just theory, companies are turning to specialized platforms like Tangibly, an AI-driven tool designed specifically for trade secret management. Tangibly helps organizations identify, catalog, and protect their trade secrets systematically, transforming intangible IP into defensible assets.
For instance, in the technical layer, Tangibly’s AI tools automate the discovery of potential trade secrets within codebases, datasets, and documents, flagging risks like unauthorized sharing. On the legal side, it supports contract audits and compliance tracking, ensuring alignment with regulations. Ethically, by promoting proactive governance, it fosters a culture of responsibility, reducing the likelihood of misuse.
In the AI governance context, Tangibly exemplifies how the layered model can be operationalized. A company developing an AI model might use Tangibly to secure proprietary training data (technical), document it for legal protection (regulatory), and ensure it doesn’t infringe on societal norms (ethical). This not only mitigates risks but also enhances value: well-managed trade secrets can boost valuations and partnerships. Recent launches like Trade Secret 2025 (TS25) provide resources for executives, including AI-powered asset cataloging and compliance features.
Why This Matters Now
As AI adoption accelerates, the layered model reminds us that governance must be adaptive and comprehensive. Relating it to trade secrets underscores a key pain point: in an era where AI IP is often the crown jewel, tools like Tangibly make governance tangible by providing actionable insights and automation. Businesses ignoring this risk falling behind, while those embracing it can lead with confidence.
Whether you’re a startup founder, a policy maker, or an AI enthusiast, consider how the layered model applies to your work. Start by assessing your trade secrets: platforms like Tangibly can be a game-changer in turning governance from a checklist into a strategic advantage.
Interested to chat through your AI governance strategy? Connect with our team of experts.
