The rapid adoption of Artificial Intelligence (AI) in business operations, particularly in human resources, presents both transformative opportunities and significant regulatory challenges. For companies operating in Illinois and utilizing AI for video interviews, understanding and complying with the Illinois AI Video Interview Act (AIVIA) is not just a best practice—it's a legal imperative. While AIVIA doesn't explicitly mandate an "AI Impact Assessment" in the same terms as newer, broader regulations like the EU AI Act or the Colorado AI Act, its stringent requirements for transparency, explainability, and bias mitigation implicitly necessitate a robust internal assessment process.

This article provides a comprehensive, step-by-step guide for conducting an AI impact assessment under AIVIA, detailing what to document and offering practical tips for ensuring ongoing compliance. Leveraging an advanced AI compliance software solution like AICompliant can significantly streamline this process, transforming complex legal obligations into manageable, automated workflows.

Understanding the Illinois AI Video Interview Act (AIVIA)

The Illinois AI Video Interview Act (AIVIA), Bill HB 2557, became effective on January 1, 2020. It stands as one of the pioneering pieces of legislation in the U.S. addressing AI in employment. AIVIA specifically targets employers who utilize AI programs to analyze video interviews for job applicants. The primary goal of the Act is to protect job applicants from potential biases and lack of transparency inherent in AI-driven hiring processes.

Key Requirements of AIVIA:

Notice and Consent: Before an applicant participates in a video interview that will be analyzed by AI, the employer must:
- Notify the applicant that AI will be used to analyze their video submission.
- Explain how the AI works and what characteristics it uses to evaluate applicants (e.g., facial expressions, tone of voice, word choice).
- Obtain the applicant's explicit consent to be evaluated by the AI.
Limited Sharing: Employers are prohibited from sharing applicant videos, except with individuals whose expertise is necessary to evaluate the applicant for the position.
Data Retention and Deletion: Employers must delete an applicant's video and all associated copies within 30 days of receiving a request from the applicant.
No Exclusive AI Use: The Act does not prohibit human review; rather, it aims to ensure transparency where AI is part of the process.

Penalties for Non-Compliance:

Non-compliance with AIVIA can result in significant financial penalties. The Act stipulates fines up to $1,000 per violation. Enforcement is handled by the Illinois Department of Commerce and Economic Opportunity. While seemingly lower than penalties under other laws like the EU AI Act (up to $35,000,000 per violation) or the upcoming Colorado AI Act (up to $20,000 per violation), accumulating multiple violations can lead to substantial financial exposure and reputational damage.

The Implicit Mandate for an AI Impact Assessment

While AIVIA does not use the explicit term "AI Impact Assessment" (AIIA) or "Algorithmic Impact Assessment" (AIA) like NYC Local Law 144 (effective July 5, 2023) for Automated Employment Decision Tools (AEDTs) or the forthcoming Colorado AI Act (SB 24-205, effective June 30, 2026), its requirement for employers to "explain how the artificial intelligence program works and what characteristics it uses to evaluate applicants" fundamentally demands an internal assessment.

To adequately explain an AI system, an employer must first deeply understand it. This understanding forms the bedrock of an effective AI risk management framework. Without thoroughly scrutinizing the AI's design, data, and decision-making process, providing a truly accurate and transparent explanation to applicants is impossible. This article frames the necessary internal due diligence required by AIVIA as an essential, practical AI impact assessment—a critical step for automated AI compliance.

Step-by-Step Guide to AIVIA AI Impact Assessment

Conducting an AI impact assessment for AIVIA involves a structured approach to understand, evaluate, and mitigate risks associated with AI-powered video interview analysis. This process helps ensure compliance and fosters ethical AI use.

1. Define Scope and Identify AI Systems

Objective: Clearly delineate which AI systems are subject to AIVIA and the specific uses cases.

Inventory AI Tools: Compile a comprehensive list of all AI programs, platforms, or services used for analyzing video interviews in the hiring process. This includes proprietary systems, third-party vendor solutions, and embedded AI features within broader HR software.
Identify Data Flows: Map out how video interview data is collected, processed by the AI, and where the results are stored or accessed.
Vendor Due Diligence: If using third-party AI, engage with vendors to understand their system's capabilities, limitations, and compliance posture. Request documentation on their AI's design, testing, and fairness evaluations.
Documentation: Record the name of the AI system, its vendor, primary function, deployment date, and the specific roles/departments using it.

2. Understand AI Functionality and Purpose

Objective: Gain a detailed understanding of how the AI system works and what it evaluates to fulfill AIVIA's explainability requirement.

Mechanism of Operation: Document the AI's core algorithms, machine learning models, and the general approach it takes to analyze video interviews. Is it natural language processing (NLP), computer vision, or a combination?
Characteristics Analyzed: Precisely identify the "characteristics" the AI uses to evaluate applicants. This could include:
- Verbal Cues: Word choice, vocabulary, coherence, speech patterns, sentiment analysis of spoken words.
- Non-Verbal Cues: Facial expressions (e.g., perceived emotions like happiness, engagement, surprise), body language, eye contact, gestures.
- Vocal Cues: Tone of voice, pitch, speaking speed, pauses.
- Important Note: The explanation provided to candidates must be specific. Simply stating "the AI analyzes your communication style" is insufficient. It needs to detail which aspects of communication style (e.g., "AI analyzes your speech rate and the sentiment of your responses").
Decision Logic/Output: How does the AI translate its analysis into actionable outputs or recommendations (e.g., a score, a rank, a flag for certain traits)?
Documentation: Create a detailed description of the AI's functionality, including the specific characteristics it analyzes and how those analyses contribute to the evaluation process. This forms the basis of your applicant explanation.

3. Identify and Assess Potential Risks and Biases

Objective: Proactively identify and evaluate potential sources of algorithmic discrimination and other adverse impacts. This is crucial for algorithmic discrimination prevention.

Bias Detection: Investigate potential biases in the AI system.
- Training Data Bias: Is the training data representative of all applicant demographics (race, gender, age, disability, etc.)? Unrepresentative data can embed and amplify existing societal biases. Referencing concerns similar to those addressed by California AB 2013 (effective January 1, 2025) regarding training data transparency and impact can be helpful, even if AB 2013 applies to specific AI models.
- Algorithmic Bias: Does the algorithm itself exhibit unfair preferences or disparate impact across different demographic groups?
- Proxy Bias: Does the AI inadvertently use proxies for protected characteristics? For example, analyzing linguistic patterns that correlate with certain accents or socio-economic backgrounds.
Fairness Metrics: Assess the AI system's performance across different demographic subgroups using relevant fairness metrics (e.g., equal accuracy, demographic parity, equal opportunity).
Explainability Limitations: Can the AI's output be easily understood and interpreted by humans? Black-box models pose greater risks for transparency and accountability.
Other Risks: Consider risks related to data security breaches, misuse of data, or unintended consequences of the AI's recommendations. The NIST AI Risk Management Framework (AI RMF 1.0) provides a robust methodology for identifying, measuring, and managing AI risks, serving as an excellent guide for this phase.
Documentation: Catalog all identified biases and risks, their potential impact (e.g., disproportionate exclusion of certain applicant groups), and their severity.

4. Evaluate Data Handling and Privacy Practices

Objective: Ensure that the collection, storage, and processing of video interview data comply with AIVIA and broader data privacy principles.

Data Collection: Is the video data collected ethically and legally? Are applicants fully aware of what data is being recorded?
Storage and Security: Where are the video files and AI analysis results stored? Are robust security measures (encryption, access controls) in place to protect this sensitive personal data? Consider standards like ISO/IEC 42001:2023 for AI management systems.
Retention Policies: Does the organization have clear data retention policies that align with AIVIA's requirements (deletion upon request within 30 days) and other relevant regulations (e.g., GDPR, if applicable to EU candidates)?
Data Access: Who has access to the raw video data and the AI-generated insights? Is access limited to authorized personnel?
Third-Party Transfers: If data is shared with vendors or other third parties, are robust data processing agreements (DPAs) in place?
Documentation: Detail data collection methods, storage locations, security protocols, retention schedules, and data sharing agreements.

5. Review Notice and Consent Mechanisms

Objective: Verify that the process for informing applicants and obtaining their consent meets AIVIA's specific requirements.

Clarity and Comprehensiveness: Is the notice provided to applicants clear, concise, and easily understandable? Does it fully explain how the AI works and all the characteristics it uses for evaluation, as identified in Step 2? Avoid legal jargon.
Timeliness: Is the notice provided before the applicant participates in the video interview?
Explicit Consent: How is consent obtained? Is it an affirmative action (e.g., checking a box) or implied? AIVIA requires explicit consent.
Opt-Out Mechanisms: Although AIVIA doesn't mandate an alternative, it's good practice to consider if an applicant can reasonably opt-out of AI analysis, perhaps by requesting a human-reviewed interview. While not a direct AIVIA requirement, it aligns with broader ethical AI principles and may become a requirement under future legislation (e.g., Colorado AI Act's emphasis on opt-outs for high-risk systems).
Documentation: Keep records of the exact notice language presented to applicants, the consent mechanism, and proof of consent for each applicant. An AI compliance checklist 2026 often includes specific items for notice and consent for various state laws.

6. Develop Mitigation Strategies

Objective: Create and implement plans to address identified risks and biases, ensuring the AI system is used fairly and transparently.

Bias Remediation:
- Data Rebalancing: If training data is biased, explore methods to collect more representative data or rebalance existing datasets.
- Algorithmic Adjustments: Implement bias-mitigation techniques within the AI algorithm (e.g., pre-processing data, in-processing constraints, post-processing outputs).
- Fairness Metrics Monitoring: Continuously monitor the AI's performance across demographic groups.
Human Oversight: Implement human review of AI-generated recommendations, especially for critical decisions. Ensure human decision-makers are empowered to override AI outputs.
Transparency Enhancements: Refine the explanation provided to applicants based on the assessment's findings. Make it even clearer and more specific.
Vendor Engagement: Work with AI vendors to address identified risks and ensure they are taking steps to improve their systems' fairness and transparency.
Policy Updates: Update internal policies and procedures to reflect responsible AI use, data handling, and AIVIA compliance.
Documentation: Outline all proposed and implemented mitigation strategies, including timelines, responsible parties, and expected outcomes.

7. Document Everything (AI Impact Assessment Template)

Objective: Create a comprehensive record of the entire assessment process, findings, and mitigation actions. This serves as vital evidence of compliance.

Centralized Repository: Consolidate all documentation in an accessible, secure location. This includes:
- AI system inventory.
- Detailed functionality descriptions.
- Bias and risk assessment reports.
- Data privacy evaluations.
- Notice and consent forms/records.
- Mitigation plans and their implementation status.
- Records of periodic reviews.
Structured Reporting: Utilize an AI impact assessment template to standardize documentation. This template should capture all required elements, making it easier to demonstrate compliance during an audit.
Audit Trail: Maintain a clear audit trail of all changes, decisions, and approvals related to the AI system and its assessment.
AICompliant's Role: An AI compliance platform like AICompliant offers dedicated modules for creating, managing, and storing these assessments. Its structured environment ensures all critical information is captured, organized, and easily retrievable. This enables a robust AI risk management framework that is auditable.

8. Implement and Monitor

Objective: Put mitigation strategies into practice and continuously monitor the AI system for ongoing compliance and performance.

Operational Integration: Embed the updated policies and procedures into daily HR and IT operations.
Training: Provide ongoing training to HR personnel and other stakeholders involved in the AI-powered hiring process on AIVIA requirements, ethical AI use, and the specific explanations to provide applicants.
Continuous Monitoring: Establish mechanisms for ongoing monitoring of the AI system's performance, fairness, and compliance with AIVIA. This includes:
- Regular checks for bias.
- Reviewing applicant feedback.
- Monitoring changes in vendor AI models.
Periodic Reassessment: Conduct periodic reassessments of the AI system (e.g., annually, or whenever there are significant changes to the AI model or its use case) to ensure continued compliance and address new risks. This aligns with best practices for an adaptive AI risk management framework.
Automated Monitoring: An AI compliance tool can provide real-time dashboards and alerts for compliance metrics, facilitating proactive management and automated AI compliance.

Leveraging AICompliant for AIVIA Compliance

Navigating the nuances of AIVIA and implementing a thorough AI impact assessment can be complex and resource-intensive. This is where a specialized AI compliance software solution becomes invaluable. AICompliant's platform is designed to automate and simplify these critical processes, ensuring that your organization not only meets AIVIA's requirements but also establishes a broader, resilient AI risk management framework.

AICompliant's features streamline every step of your AIVIA impact assessment:

Structured Assessment Workflows: Our platform provides pre-built templates and guided workflows that walk you through each step of the impact assessment, ensuring no requirement is missed.
AI System Inventory & Documentation: Easily inventory all AI tools, document their functionality, characteristics analyzed, and vendor information, creating a comprehensive and auditable record in your AICompliant dashboard.
Bias & Risk Identification: Utilize built-in tools and prompts to identify potential sources of bias in training data and algorithmic design, helping you develop effective algorithmic discrimination prevention strategies.
Automated Notice & Consent Management: Generate compliant notice language, manage consent collection, and maintain an immutable record of applicant consent, simplifying this critical AIVIA requirement.
Policy & Procedure Management: Centralize and manage your internal policies, ensuring they reflect AIVIA's requirements and are easily accessible for staff training.
Audit-Ready Documentation: AICompliant automatically compiles all assessment data, mitigation plans, and compliance artifacts into a single, auditable repository, ready for review by the Illinois Department of Commerce and Economic Opportunity or internal stakeholders. You can quickly generate an AI compliance checklist 2026 report for all your active regulations.
Continuous Monitoring & Alerts: Set up automated reminders for periodic reviews and receive alerts for potential compliance gaps, fostering automated AI compliance and proactive risk management.
US State AI Law Tracker Integration: Our platform also helps you stay abreast of the evolving US state AI law tracker landscape, including upcoming regulations like the Colorado AI Act (effective June 30, 2026) and California SB 942 (effective January 1, 2026), allowing you to scale your compliance efforts efficiently.

For an initial assessment of your current AI compliance posture, explore our compliance checker tool.

Beyond AIVIA: A Proactive Approach to AI Governance

While AIVIA focuses specifically on video interviews, it serves as a critical example of the broader trend in AI regulation. Companies must adopt a proactive, holistic approach to AI governance. Laws like NYC Local Law 144 (effective July 5, 2023), the forthcoming Colorado AI Act (effective June 30, 2026), and the EU AI Act (enforcement for high-risk systems beginning August 2, 2026) demonstrate a growing global consensus on the need for responsible AI development and deployment. These regulations introduce more explicit requirements for AI impact assessments, risk management, human oversight, and detailed documentation, often with substantially higher penalties. For instance, the EU AI Act carries penalties up to $35,000,000 per violation.

By establishing a robust AI risk management framework and utilizing an AI compliance platform now for specific laws like AIVIA, your organization will be well-positioned to adapt to the accelerating pace of global AI legislation. Building a culture of ethical and compliant AI use is not merely about avoiding penalties; it's about building trust with applicants, employees, and customers, and safeguarding your organization's reputation in the AI era.

Conclusion

The Illinois AI Video Interview Act (AIVIA) is a foundational regulation requiring employers to be transparent and accountable when using AI in video interviews. By treating AIVIA's requirements for explanation and bias mitigation as an implicit mandate for an AI impact assessment, organizations can systematically identify, evaluate, and address potential risks. This proactive approach, supported by robust AI compliance software like AICompliant, ensures not only legal adherence but also the ethical and fair treatment of all job applicants. Establishing a strong AI risk management framework today is paramount for navigating the complex AI regulatory landscape of tomorrow.

Get Started with AICompliant

Ensure your organization is fully compliant with the Illinois AI Video Interview Act and future AI regulations. Discover how AICompliant's platform can streamline your AI impact assessments, automate compliance, and safeguard your operations.

Explore AICompliant Pricing & Solutions

Frequently Asked Questions

Does the Illinois AI Video Interview Act (AIVIA) explicitly require an "AI Impact Assessment"?

AIVIA (Bill HB 2557, effective January 1, 2020) does not explicitly use the term "AI Impact Assessment." However, it requires employers to explain "how the artificial intelligence program works and what characteristics it uses to evaluate applicants." To provide this explanation accurately and identify potential biases, organizations must conduct an internal assessment that effectively functions as an AI impact assessment, understanding the AI's functionality and risks.

What are the penalties for violating the Illinois AI Video Interview Act?

Non-compliance with AIVIA can result in penalties of up to $1,000 per violation. Enforcement is carried out by the Illinois Department of Commerce and Economic Opportunity. While this may seem modest compared to other regulations like the EU AI Act (up to $35,000,000 per violation), repeated violations can quickly accumulate to substantial financial risk and reputational damage.

How can an AI compliance software platform help with AIVIA compliance?

An AI compliance software platform like AICompliant can significantly streamline AIVIA compliance by providing structured workflows for documenting AI systems, managing notice and consent, conducting bias assessments, and maintaining an auditable record of all compliance activities. It helps automate the creation of an AI impact assessment template and ensures ongoing adherence to the Act's requirements, serving as an AI compliance tool for automated AI compliance.

Is AIVIA the only state-level AI regulation I need to worry about in the US?

No, AIVIA is just one example of a rapidly growing number of state and local AI regulations in the US. Other notable laws include NYC Local Law 144 (Automated Employment Decision Tools, effective July 5, 2023), the Colorado AI Act (SB 24-205, effective June 30, 2026), and various data privacy laws with AI implications, such as the Virginia CDPA (effective January 1, 2023) and California SB 942 (effective January 1, 2026) for AI transparency. Organizations need a comprehensive US state AI law tracker and robust AI risk management framework to navigate this evolving landscape.

What specific information about the AI should be provided to applicants under AIVIA?

Under AIVIA, employers must explain "how the artificial intelligence program works and what characteristics it uses to evaluate applicants." This means detailing the types of analyses performed (e.g., natural language processing for word choice, computer vision for facial expressions, sentiment analysis) and the specific traits or cues the AI assesses (e.g., "AI analyzes your speech rate, tone of voice, and the sentiment conveyed in your responses"). This explanation should be clear, concise, and understandable to a layperson, avoiding overly technical jargon.

AI Compliance: Illinois AIVIA Impact Assessment