Exam content outlines
Curriculum, domains, and objectives
Certified Law Enforcement AI Specialist (CLE-AIS)
75 questions · 90 minutes · 80% to pass
AI Foundations for Public Safety
15% of the exam
Understand how modern AI systems work, what their outputs mean, and where they fail, at the level a practitioner needs to use and question them.
- 1.1Define artificial intelligence, machine learning, deep learning, generative AI, and large language models (LLMs) at a practitioner level.
- 1.2Explain training data, models, and inference, and why AI outputs are probabilistic rather than verified facts.
- 1.3Recognize hallucinations, fabricated citations, and overconfident outputs, and apply verification strategies.
- 1.4Distinguish predictive analytics, generative AI, and computer-vision systems used in policing.
- 1.5Explain prompts, context windows, and retrieval-augmented tools at a basic level.
- 1.6Interpret accuracy concepts: false positives and negatives, thresholds, confidence scores, and base rates.
- 1.7Recognize privacy implications of vendor data retention and use of agency data for model training.
- 1.8Identify inherent limits of AI: no true understanding, model drift, and bias inherited from data.
Synthetic Media & Evidence Authentication
20% of the exam
Recognize deepfakes and synthetic media, authenticate digital media offered as evidence, and address authenticity disputes in court.
- 2.1Recognize deepfake video, face swaps, lip-sync manipulation, voice cloning, and AI-generated images.
- 2.2Explain why visual or audio 'tells' and automated deepfake detectors are unreliable as sole proof.
- 2.3Authenticate media by obtaining originals from the source, preserving metadata, hashing, and maintaining chain of custody.
- 2.4Explain content provenance (C2PA Content Credentials) and watermarking, including their limits.
- 2.5Corroborate media with independent evidence such as witnesses, CCTV, device data, and timestamps.
- 2.6Understand courtroom issues: authentication, expert testimony, and the 'liar's dividend.'
- 2.7Respond when genuine evidence is claimed to be fake, or fabricated media is presented as real.
- 2.8Document analysis steps and engage qualified forensic examiners when needed.
AI-Enabled Crime & Investigations
20% of the exam
Investigate crimes that use AI as a tool, support victims, and preserve evidence from AI platforms.
- 3.1Investigate voice-clone 'family emergency' and impersonation scams.
- 3.2Recognize deepfake executive impersonation, business email compromise, and video-call fraud.
- 3.3Respond to AI-generated child sexual abuse material: illegality, NCMEC CyberTipline reporting, and victim support.
- 3.4Investigate sextortion and non-consensual intimate imagery created or manipulated with AI.
- 3.5Recognize AI-assisted phishing, malware, and social engineering at scale.
- 3.6Investigate synthetic identities and AI-generated documents.
- 3.7Apply victim-centered response and prevention messaging (family code words, call-back verification).
- 3.8Preserve evidence held by AI platforms and obtain records through legal process.
Responsible & CJIS-Compliant Use of AI Tools
15% of the exam
Use AI tools in agency work lawfully and safely: protecting CJI, keeping humans accountable, and managing records and discovery.
- 4.1Never enter CJI or PII into public or unapproved AI tools; identify agency-approved tools.
- 4.2Evaluate AI vendor agreements: CJIS Security Addendum, data retention, and training on agency data.
- 4.3Use AI-assisted report writing responsibly: officer review, accuracy, and personal accountability.
- 4.4Understand disclosure, discovery, and public-records implications of AI use, including prompts and drafts.
- 4.5Use AI transcription and translation tools with human verification.
- 4.6Recognize prompt injection, data leakage, and other security risks of AI tools.
- 4.7Distinguish appropriate uses from inappropriate ones (e.g., charging decisions, credibility determinations).
- 4.8Maintain meaningful human review for consequential decisions.
Biometrics & AI-Powered Surveillance
15% of the exam
Use facial recognition, ALPR, gunshot detection, and video analytics within their limits, with documentation and policy authority.
- 5.1Treat facial recognition results as investigative leads only, never as sole probable cause; require human review.
- 5.2Explain probe image quality, candidate lists, thresholds, and demographic differentials in face recognition.
- 5.3Document facial recognition use, policy authorization, and disclosure obligations.
- 5.4Verify ALPR hits before acting, and follow data retention and sharing limits.
- 5.5Treat gunshot detection and video analytics alerts as leads, accounting for accuracy limits.
- 5.6Recognize privacy considerations for AI-enabled drones, real-time crime centers, and camera networks.
- 5.7Apply lessons from misidentifications and wrongful arrests linked to over-reliance on technology.
- 5.8Recognize that state and local laws on biometric technology vary; follow policy and legal counsel.
AI Governance, Bias & Policy
15% of the exam
Govern AI responsibly: identify bias, manage risk, write policy, and maintain community trust.
- 6.1Identify sources of algorithmic bias and their impact on communities.
- 6.2Recognize automation bias and confirmation bias in human-AI decision making.
- 6.3Explain predictive policing risks, including feedback loops from historical data.
- 6.4Apply the NIST AI Risk Management Framework functions (Govern, Map, Measure, Manage).
- 6.5Build an agency AI policy: inventory, approval process, permitted and prohibited uses, and training.
- 6.6Audit, monitor, and evaluate AI systems over time, including incident reporting.
- 6.7Promote transparency and community engagement, including public reporting on AI use.
- 6.8Conduct procurement due diligence: test vendor claims and seek independent validation.
Study with practice quizzes and an AI coach
Free with every candidate account. Practice items are separate from the live exam banks.