Infinite Campus AI-Driven Student Risk Model Leads to Cuts in Support for Nevada's Low-Income Schools

An AI system developed by Infinite Campus and deployed by Nevada to identify at-risk students led to a sharp reduction in the number classified as needing support, dropping from 270,000 to 65,000. The reclassification caused significant budget cuts in schools serving low-income populations. The drastic reduction in identified at-risk students reportedly left thousands of vulnerable children without resources and support.

Risk classification

• Primary risk domain: 1 Discrimination & Toxicity
• Primary risk subdomain: 1.1 Unfair discrimination and misrepresentation

The AI system initially utilized sensitive demographic characteristics like gender, race, and birth country, leading to concerns of unfair discrimination where students with identical academic profiles received different risk scores based on gender.

Additional risk subdomains

• 7.4 Lack of transparency or interpretability: The algorithm's specific weights and decision-making processes were kept proprietary by Infinite Campus, leaving educators frustrated by the 'black box' system.
• 7.3 Lack of capability or robustness: The system failed to robustly identify clearly vulnerable students, such as homeless and low-income children, as being at risk.

Causal factors

• Entity: AI
• Intent: Unintentional
• Timing: Post-deployment

The budget shortfalls and exclusion of vulnerable students were unexpected outcomes resulting from the deployment of the AI model to optimize funding allocation.

EU AI Act risk tier

• Risk tier: 2 High Risk

Risk Level 2: High Risk. The system is an educational tool used to determine student risk levels and allocate funding, which directly affects access to education. The EU AI Act classifies 'Educational and vocational training systems affecting access to education' as High Risk.

AI system and alleged parties

• AI system: Infinite Campus AI-Driven Student Risk Model (Infinite Campus)
• AI purpose: Resource Allocation; Identification
• Behaviour type: Assistant
• Alleged developer: Infinite Campus
• Alleged deployer: Nevada Department of Education
• Alleged harmed parties: Somerset Academy, Nevada school districts, Mater Academy of Nevada, Low-income students in Nevada

Harm severity

Highest direct severity in any category: Substantial. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.

• Physical: direct Negligible, indirect Negligible
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Substantial, indirect Substantial
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Negligible, indirect Negligible
• Differential treatment: direct Substantial, indirect Substantial
• Civil rights: direct Negligible, indirect Negligible
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Negligible, indirect Negligible
• Psychological: direct Negligible, indirect Negligible
• Epistemic: direct Negligible, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Financial

Reported: Yes, the report describes schools losing state funding they relied on, forcing districts to slash programs and redo budgets due to the drastic reduction in classified students.

Directly caused: The AI system's classification directly reduced the number of funded at-risk students in Nevada by over 200,000, causing immediate budget re-evaluations and program cuts across school districts.

Indirectly caused: Schools had to slash educational programs, tutoring, and support services due to the sudden loss of expected state funding.

Inferred additional harm: It is likely that the reduction in funding led to millions of dollars in budget cuts across the state's school districts, directly impacting the resources available for vulnerable students.

Differential treatment

Reported: Yes, the report describes how the AI system treated students differently based on demographic factors like gender, race, and birth country in its first year.

Directly caused: The AI model initially used gender, race, and birth country as factors, resulting in girls being classified as lower risk than boys with identical academic and behavioral profiles.

Indirectly caused: Low-income and homeless students at certain schools (like Somerset Academy and Mater Academy) were completely excluded from at-risk funding despite their clear socioeconomic vulnerabilities.

Inferred additional harm: Other vulnerable demographic groups may have been systematically under-represented and denied funding due to the proprietary weights assigned to variables like home language.

People affected

• Occurrences reported: 1
• People reportedly harmed: 200000
• People reportedly exposed: 270000

Potential causes

Management

• Prioritizing Cost Efficiency: Management sought to target dwindling funds without assessing negative impacts.
• Inadequate Risk Assessment: Decision-makers failed to foresee the impact of dropping 200,000 students.

Technology

• Proprietary Black Box Model: The algorithm workings were kept private as proprietary intellectual property.
• Narrow Definition of Risk: Model focused solely on graduation rates, ignoring mental health and wellbeing.
• Algorithmic Bias in Variables: First-year model used sensitive factors like race, gender, and birth country.

Data Inputs

• Proxy Metrics for Engagement: Using guardian portal logins as a proxy for student risk is flawed.
• Incomplete Risk Indicators: Data inputs lacked metrics on student depression, self-harm, and food security.

Human Factors

• Overreliance on Predictive Tech: State administrators trusted the model output without local verification.
• Lack of Technical Literacy: School leaders did not understand how the automated scoring system worked.

Process and Methods

• Lack of Transparency: No clear explanation was provided to schools on how scores were calculated.
• No Validation Mechanisms: The state implemented the system without independent auditing or evaluation.
• Abrupt Funding Reallocation: Budgets were slashed suddenly based on model outputs without transition plans.

Regulatory Environment

• No Algorithmic Oversight: There were no state regulations governing the use of AI in school funding.
• Absence of Standardized Audits: No regulatory requirement existed to audit the predictive model for bias.

Information quality

• Classification confidence: High
• Reason for confidence: The report provides clear details on the AI developer (Infinite Campus), the specific metrics used, the scale of the impact (reducing at-risk students from 270,000 to 65,000), and the resulting funding issues for specific schools. The role of the AI is explicitly defined.
• Ambiguities identified: The exact proprietary weights of the algorithm's variables are kept private, making it a 'black box' to educators.
• Alternative interpretations: The funding cuts could be viewed as a policy decision by the state of Nevada on how to define 'at-risk' rather than purely an AI failure, though the AI's specific classifications directly drove the outcome.

The deployment of an AI-based student risk model in Nevada led to significant, biased reductions in educational funding for over 200,000 vulnerable students. While causing notable domestic policy challenges and highlighting risks in algorithmic governance, the incident presents minimal threats to core national security.

• Overall national security impact: Minor
• Response level: Moderate
• Scope: Single nation
• Primary target: United States
• Alleged perpetrator: Unknown

Threat characteristics

• Imminence: Long-term. Represents an ongoing domestic policy and algorithmic governance concern rather than an active, imminent national security crisis.
• Autonomy: Human-supervised. The AI acted as an assistant to classify students, but state administrators ultimately executed the funding cuts based on its outputs.
• Novelty: Established threat. Algorithmic bias and transparency issues in public resource allocation are well-established challenges in AI governance.

Impact by dimension

• Physical security: Negligible. No physical threat, kinetic attacks, or critical infrastructure disruption occurred in this educational funding incident.
• Information security: Negligible. No intelligence compromise, espionage, or systematic information warfare operations were associated with this incident.
• Sovereignty: Minor. Disrupted state-level educational funding allocation and school budgeting processes, representing a minor impact on local government administration.
• Economic security: Negligible. While local school budgets were severely impacted, there was no threat to national economic stability or strategic technological competitive advantage.
• Societal stability: Minor. The algorithm initially incorporated sensitive demographic criteria and systematically reduced funding eligibility for over 200,000 vulnerable and low-income students.