TinyML Security Training is a focused two-day program designed to help professionals understand the security risks, threats, and protection strategies associated with deploying machine learning models on ultra-low-power and resource-constrained devices. This course covers security considerations across the TinyML lifecycle, including data collection, model deployment, device integrity, and operational resilience, enabling participants to design and manage secure, trustworthy, and resilient TinyML solutions for real-world edge environments.

• Understand TinyML security fundamentals
• Learn secure embedded AI concepts
• Understand edge device security basics
• Gain knowledge of AI threat detection
• Learn secure TinyML deployment practices
• Understand data privacy in TinyML systems
• Explore IoT security concepts
• Identify TinyML security use cases

• Embedded systems and IoT engineers
• TinyML and edge AI practitioners
• Security engineers working with edge devices
• Product developers building smart and connected devices
• Technology professionals responsible for device security

• Basic understanding of TinyML, edge AI, or embedded systems concepts
• Familiarity with IoT devices, sensors, or microcontrollers is beneficial
• Awareness of basic cybersecurity or system security concepts
• No advanced cryptography or hardware security background required

• Understanding security risks specific to TinyML deployments
• Identifying attack surfaces in TinyML systems
• Applying security controls for models and devices
• Protecting data, models, and inference pipelines
• Implementing responsible and resilient TinyML security practices

By the end of this training, participants will be able to identify and assess security risks in TinyML systems, understand common attack vectors, apply appropriate security controls to protect data and models, and support secure, resilient, and responsible deployment of TinyML solutions on resource-constrained devices.

Introduction to TinyML Security
• Why security is critical for TinyML systems
• Differences between cloud AI security and TinyML security
• Threat landscape for edge and microcontroller-based AI
• Security responsibilities across the TinyML lifecycle

TinyML Attack Surfaces and Threat Models
• Physical access and device tampering risks
• Model extraction and intellectual property theft
• Data poisoning and adversarial sensor inputs
• Side-channel and inference-based attacks

Secure Data and Model Handling
• Protecting training and inference data
• Model integrity and authenticity checks
• Secure storage of models on devices
• Managing updates and version control securely

Device-Level Security for TinyML
• Secure boot and firmware protection concepts
• Hardware-based security features overview
• Preventing unauthorized access and modification
• Managing secrets and credentials on devices

Operational Security and Resilience
• Monitoring device behavior and anomalies
• Handling failures and compromised devices
• Secure deployment and lifecycle management
• Balancing security with power and performance constraints

Responsible and Secure TinyML Deployment
• Privacy considerations for on-device inference
• Ethical risks and misuse prevention
• Security governance for large-scale TinyML deployments
• Best practices for secure and sustainable TinyML systems

• TinyML security fundamentals
• Embedded AI security concepts
• Edge device protection techniques
• IoT security basics
• Secure model deployment workflows
• Data privacy and protection
• AI threat detection concepts
• Access control and authentication basics
• Security risk management concepts
• Practical TinyML security scenarios

• TinyML security threat analysis exercise
• Device and model protection scenario discussion
• Secure deployment assessment activity
• Final knowledge evaluation quiz

Participants will receive course materials, slides, reference materials, exercises and access to resources for further learning.

Participants who successfully complete the training will receive an AcadNXT Certification in TinyML Security Training, validating their expertise in identifying threats, securing models and devices, and applying best practices for secure TinyML deployment.