AI Usage Policy

Introduction

Tactical Edge AI (“Tactical Edge,” “we,” “us,” or “our”) builds and deploys artificial intelligence systems for enterprise clients. AI is central to our products, services, and consulting engagements.

This AI Usage Policy outlines how we develop, deploy, and govern AI systems across our organization. It reflects our commitment to building AI that is transparent, fair, accountable, and aligned with the interests of our customers, their users, and the broader public.

How We Use AI in Products and Services

AI powers the core capabilities of our products and consulting engagements, including:

• Agentic AI systems that reason, decide, and act autonomously within defined enterprise workflows
• Large Language Model (LLM) integrations for natural language understanding, generation, and summarization
• Predictive analytics and recommendation engines across sales, operations, and customer engagement
• Intelligent document processing and knowledge extraction
• Voice AI agents for contact center automation and customer interaction
• Custom model fine-tuning and domain-specific model development for regulated industries

Responsible AI Principles

These five principles guide how we scope, design, and operate AI systems. Specific controls are selected according to the use case, risk assessment, customer agreement, and applicable law:

• Transparency: We clearly communicate when AI is being used, what data it processes, and how decisions are made. Users and stakeholders can understand the role AI plays in any interaction or outcome.
• Fairness: We design and test AI systems to avoid unjust bias and discrimination. Outputs are evaluated for equitable treatment across demographic groups and use cases.
• Accountability: Every AI system has a designated owner responsible for its behavior, performance, and compliance. We maintain clear audit trails and escalation paths.
• Privacy: AI systems are designed with privacy by default. We minimize data collection, enforce access controls, and comply with applicable data protection regulations.
• Safety: We evaluate AI systems for potential harms before deployment and implement safeguards, guardrails, and fallback mechanisms to prevent unintended consequences.

Human Oversight and Governance

AI systems at Tactical Edge operate under structured human oversight:

• Risk assessments identify consequential actions that require human review or approval before execution
• Autonomous agents operate within clearly defined permission boundaries and escalation policies
• Designated governance and delivery owners review new AI deployments, evaluate risk, and document production approval responsibilities
• Model and workflow performance is reviewed on a cadence appropriate to risk and contracted operations
• Systems that can take consequential actions include scope-appropriate suspension, rollback, or containment mechanisms

Data Handling in AI Systems

Data used in our AI systems is handled with strict controls:

• Customer data is not used to train shared or general-purpose models. Customer-specific training or fine-tuning requires written customer instructions, confirmation of the necessary data rights, and governing contract terms
• When training or fine-tuning is contracted, approved datasets undergo scope-appropriate quality, provenance, relevance, privacy, and bias review before use
• Personal data is minimized and, where appropriate to the use case, redacted, tokenized, anonymized, or pseudonymized in AI pipelines
• Training, inference, prompt, output, and operational-log retention follows the applicable product policy, customer agreement, data processing terms, and configured controls
• Access to AI training data and model artifacts is restricted by role-based access controls

Model Selection and Evaluation

We select and evaluate AI models based on rigorous criteria:

• Task-specific performance benchmarks and accuracy thresholds
• Latency, cost, and scalability requirements for production workloads
• Security posture and data handling practices of model providers
• Licensing terms, intellectual property considerations, and compliance with customer contractual obligations
• Interpretability and explainability of model outputs for the intended use case
• Ongoing monitoring of model drift, degradation, and emergent behaviors in production

Bias Monitoring and Mitigation

We actively work to identify and reduce bias in our AI systems:

• Risk assessments determine when pre-deployment bias testing is required for approved data and model outputs
• Where fairness risks are relevant, metrics and review criteria are defined for the use case and affected populations
• Feedback mechanisms allow users and stakeholders to report potential bias or harmful outputs
• Identified biases trigger remediation workflows including data rebalancing, prompt engineering adjustments, or model retraining
• Third-party audits may be engaged for high-risk AI applications

Customer Data Protection in AI Pipelines

Protecting customer data within AI processing pipelines is a non-negotiable commitment:

• Encryption in transit and at rest is configured for applicable customer data stores, services, and network paths in the approved architecture
• Customer-data isolation follows the selected product, cloud-account, tenant, and service architecture
• Inference logs containing sensitive data are subject to retention limits and access controls
• Prompt, output, review, and workflow records are retained only as described by the applicable product configuration, customer agreement, retention policy, and legal requirements
• Data processing agreements and service-specific terms govern AI-related data handling where applicable

Compliance with Emerging AI Regulations

We monitor and adapt to the evolving global regulatory landscape for AI, including:

• The EU AI Act and its risk-based classification framework
• US federal and state AI governance requirements, including executive orders and sector-specific guidance
• International standards such as ISO/IEC 42001 (AI Management Systems) and NIST AI Risk Management Framework
• Industry-specific regulations for AI in healthcare, finance, and government sectors

Our governance processes are designed to be adaptable so that compliance obligations can be met as regulations are enacted and evolve.