AI Productivity Creates New Cybersecurity Risks—Learn How to Control Them
Identify the hidden risks in prompts, uploads, connectors, shared workspaces, browser extensions, and AI-enabled agents.
Use AI with greater confidence while protecting the information, systems, and professional responsibilities entrusted to you. Enroll in Cybersecurity and Confidentiality Risks of AI Tools in Engineering.
Artificial intelligence can help engineers work faster, but routine actions such as pasting text into a prompt, uploading a specification, connecting a project drive, or executing generated code can expose confidential information or create new cybersecurity vulnerabilities.
This course provides a practical framework for identifying and controlling those risks. Learners will examine data classification, tool authorization, account settings, retention, prompt injection, secure output handling, access control, vendor review, incident response, and professional accountability.
Course Syllabus
Module 1 — Why AI Changes the Engineering Security Boundary
Understand how prompts, files, integrations, and connected tools expand the systems that must be protected.
Module 2 — Data Classification and Confidentiality Decisions
Learn how to identify sensitive engineering information and determine what may be entered into an AI system.
Module 3 — Accounts, Vendors, Retention, and Model Training
Evaluate AI account types, data retention, provider practices, subprocessors, and changing platform features.
Module 4 — Prompts, Uploads, Connectors, and Information Leakage
Recognize how chat history, documents, metadata, memory, shared links, and integrations can expose project information.
Module 5 — Prompt Injection and Untrusted Content
Understand how malicious instructions embedded in documents, websites, email, images, or code can manipulate AI behavior.
Module 6 — Secure Handling of AI Outputs, Code, and Calculations
Apply review, testing, verification, and sanitization practices before using AI-generated technical content.
Module 7 — Identity, Access, Integrations, and Excessive Agency
Limit permissions, secure accounts, protect credentials, and retain human approval over consequential actions.
Module 8 — Governance, Documentation, and Vendor Due Diligence
Develop practical policies, records, approval criteria, and organizational controls for AI-assisted work.
Module 9 — Incident Response for AI-Related Disclosures
Learn how to recognize, contain, document, and respond to suspected data exposure or unsafe AI activity.
Module 10 — Practical Engineering Workflow and Final Review
Apply a repeatable process to define, classify, authorize, minimize, restrict, verify, document, and approve AI-assisted work.