Graduate (10-12 weeks) Intern - AI/ML for Cybersecurity

National Laboratory of the Rockies • Golden, CO

7d • Onsite

About The Position

The intern will conduct hands-on research at the intersection of machine learning, statistics, and critical energy infrastructure security. The role focuses on enhancing the reliability, transparency, and operator-readiness of AI-driven cybersecurity tools used to secure power grid substations and physical assets, with an emphasis on explainability and robust uncertainty bounds.

The intern will design and implement explainable AI methods, apply statistical techniques to quantify model confidence and predictive uncertainty, and analyze cyber-physical power system security datasets to validate anomaly-detection behavior. The work may also explore machine-learning safety techniques to improve model behavior under distribution shifts or input corruptions, and support visualization and alert design so operators can understand explanations and confidence metrics when prioritizing alerts.

The intern will collaborate with cybersecurity researchers and AI development teams to integrate explainability modules, document results, and develop guidance for deploying reliable AI models. The internship is listed as 10–12 weeks, fixed-term, and 40 hours per week.

Requirements

• Minimum cumulative GPA of 3.0.
• Meet the applicable student or recent-graduate eligibility requirements: full-time enrollment in an accredited bachelor's or master's program; completion of a bachelor's or master's degree within the past 12 months; or completion of a master's degree while enrolled in an accredited Ph.D. program. Educational requirements must be met before the employment start date.
• Currently pursuing or completed a bachelor's, master's, or Ph.D. in Computer Science, Cybersecurity, Electrical Engineering, AI, Applied Mathematics, or Statistics.
• Strong proficiency in Python.
• Hands-on experience with PyTorch, TensorFlow, or Hugging Face.
• Substantial coursework or project experience in explainable AI, machine-learning safety, confidence calibration, or statistical uncertainty modeling.
• Practical experience working with large language models, foundation models, or other complex deep neural networks.
• Submit official or unofficial school transcripts as part of the application; a letter of recommendation will be required if selected.
• Be able to obtain and maintain a federal Personal Identity Verification (PIV) card when required for an assignment extending beyond six months, including a favorable background investigation.
• Pass the required pre-employment drug test.

Nice To Haves

• Active research or publications at top-tier conferences such as NeurIPS, ICML, ICLR, AAAI, IEEE S&P, USENIX Security, or ACM CCS, focused on explainable AI, robust machine learning, or cyber-physical energy systems.
• Experience with explainability libraries such as Captum, SHAP, or LIME, and uncertainty-calibration methods.
• Familiarity with multi-sensor data fusion, power-systems modeling, or grid anomaly detection.
• Contributions to open-source software libraries focused on AI transparency, reliability, or security.

Responsibilities

• Design and implement explainable AI methods, such as attribution maps, rule extraction, and root-cause analyses, to make machine-learning models transparent and auditable for control-room operators and system administrators.
• Apply statistical techniques and frameworks, such as conformal prediction, Bayesian neural networks, or ensemble methods, to quantify model confidence and predictive uncertainty in real time.
• Analyze cyber-physical power system security datasets, including 9-Bus system baseline and cyber-physical anomaly runs, to identify feature patterns and validate anomaly-detection model behavior.
• Explore machine-learning safety techniques, including model steering, to help ensure models behave predictably under distribution shifts or input corruptions.
• Support research into visualization methods and alert design that clearly present explainability and confidence metrics to operators and assist with alert prioritization.
• Collaborate with cybersecurity researchers and AI development teams to integrate explainability modules, document results, and compile guidelines for deploying reliable AI models.

Benefits

• Medical insurance
• Dental insurance
• Vision insurance
• 403(b) Employee Savings Plan with employer match
• Sick leave where required by law
• Some positions may be eligible for relocation expense reimbursement
• Potential eligibility for performance-, merit-, and achievement-based awards with a monetary component

What This Job Offers

• Job Type: Full-time
• Career Level: Intern