AI Ethics & Bias Audit - New Orleans Bylaw Guide
In New Orleans, Louisiana, public agencies and contractors using automated decision systems must follow municipal procurement, data governance, and transparency rules when conducting AI ethics and bias audits. This guide explains where local authority lies, how audits can be structured to meet city expectations, what enforcement pathways exist, and how to report problems to city offices. It is written for municipal staff, vendors, auditors, and community advocates seeking concrete steps to reduce algorithmic harm in city programs and services while preserving public accountability.
Penalties & Enforcement
The City of New Orleans does not publish a single titled "AI ethics" ordinance on the municipal code pages; specific fine amounts and prescribed penalties for AI audit failures are not specified on the cited page[1]. Enforcement depends on the controlling instrument cited in procurement contracts, data-sharing agreements, or department rules. Typical enforcement tools available under municipal authority include fines, contract remedies, orders to cease using specific systems, injunctive or judicial remedies, and debarment or contract termination.
- Fines and monetary penalties: not specified on the cited page; amounts vary by contract or ordinance enforcement.
- Escalation: first, repeat, and continuing offence procedures are determined by the underlying contract or code section and are not specified on the cited page.
- Non-monetary sanctions: orders to stop use, system modifications, audits, suspension of procurement eligibility, and referral to courts or administrative hearings.
- Enforcer and complaint pathways: enforcing departments typically include Procurement, Department of Innovation and Technology (DoIT), the Mayor’s Office, and any department contracting the system; complaints may be submitted via the City 311 portal or department contact pages.
- Appeals and review: appeal routes follow administrative or contract dispute procedures; time limits and processes are set in the controlling procurement documents or ordinance and are not specified on the cited page.
- Defences and discretion: defences commonly include compliance with certified policies, existing permitted variances, or a demonstrated reasonable mitigation plan; specific statutory defences are not specified on the cited page.
Applications & Forms
There is no single published city form titled for "AI ethics audit" available on the municipal code page; required forms or templates are typically part of procurement or vendor compliance packets and are not specified on the cited page[1]. Municipal teams commonly require deliverables such as an audit report, data inventory, impact assessment, and corrective action plan submitted to the contracting department.
Conducting an AI Ethics & Bias Audit in New Orleans
Recommended audit scope for city systems includes governance, data quality, model testing, outcomes monitoring, vendor documentation, and public transparency. Align audit steps with procurement contract requirements and any department-specific guidance.
- Scope: systems, data flows, decision points, intended and actual impacts.
- Data review: source, completeness, representativeness, and documented consent where required.
- Technical testing: bias metrics, fairness by subgroup, performance drift, and reproducibility checks.
- Documentation: vendor model cards, training data descriptions, and algorithmic impact assessments.
- Monitoring plan: periodic reassessments, incident reporting, and thresholds for remediation.
Action Steps
- Inventory all automated decision systems used for city services and identify accountable departments.
- Commission an independent or internal audit scoped to the system’s risk level.
- Require vendors to provide model documentation and remediation plans in contracts.
- If noncompliance is found, follow contract remedies and report to the enforcing department or via 311.
FAQ
- Who enforces AI ethics and bias requirements for city systems?
- Enforcement typically sits with the contracting department, Procurement, DoIT, and the Mayor’s Office; specific enforcement depends on the contract or ordinance language.
- Are there set fines for AI audit failures in New Orleans?
- Monetary penalties are not listed on the city code page; fines and penalties depend on the controlling instrument and are not specified on the cited page[1].
- How do I report an algorithmic decision problem affecting services?
- Report service impacts through City 311 or the responsible department contact; include system name, decision instance, and supporting data.
How-To
Steps a municipal team should follow to complete a baseline AI ethics and bias audit.
- Define scope: list systems, stakeholders, and legal/regulatory constraints.
- Collect documentation: contracts, model descriptions, data schemas, and vendor disclosures.
- Run technical tests: evaluate fairness metrics, error rates by subgroup, and validation results.
- Prepare report: findings, severity ratings, recommended remediations, and timelines.
- Submit to the contracting department and track remediation until closed or escalated.
Key Takeaways
- New Orleans enforces AI issues through contracts and department rules rather than a single AI statute.
- Audits should focus on data quality, model testing, documentation, and remediation plans.
- Use 311 and department contacts to report harms and follow procurement remedies for enforcement.
Help and Support / Resources
- City of New Orleans 311 - Report an issue or request information
- City of New Orleans - Code of Ordinances
- Mayor's Office - City of New Orleans
- Department of Innovation and Technology - City of New Orleans