Artificial intelligence has already begun making decisions that shape people’s lives — who gets hired, who gets housing, who receives healthcare, credit, or public benefits, and who is flagged for investigation by law enforcement. When these systems are opaque, untested, or unaccountable, they can replicate bias, magnify harm at scale, and strip people of meaningful recourse.
AI systems trained on flawed data or deployed without safeguards have already exacerbated harm to immigrants, low-income residents, seniors, and marginalized communities. Our government has seriously lagged behind technological innovation and must set clear, enforceable rules that ensure AI strengthens human judgment, protects civil liberties, and serves the public interest.
Strong Data Protection
Limit data collection and retention by government agencies and contractors to what is strictly necessary for a defined public purpose
Require baseline cybersecurity and privacy safeguards for any entity handling state or municipal data
Impose statutory penalties for negligent data misuse, unauthorized sharing, or preventable breaches
Classify sensitive personal data held by the state as protected civic infrastructure
Require equity impact assessments for data systems that disproportionately affect vulnerable communities
Create a legal “right to be forgotten” similar to the EU
Regulate AI in High-Risk Decisions
Require human review of AI assisted decision-making in high-stakes areas like lay-offs, housing, loan approval, court sentencing, etc.
Require transparency disclosures when AI is used in hiring, housing, credit, healthcare, or public benefits decisions
Mandate bias testing and performance audits for AI systems used in consequential determinations
Require clear documentation of decision logic and system limitations
Guarantee a right to explanation and appeal for individuals affected by automated decisions
Risk-Based AI Regulation
Establish a tiered regulatory framework that applies stricter oversight to higher-risk AI uses and exempt low-risk uses from unnecessary regulatory burden to spur innovation
Regularly update risk classifications to reflect evolving technology and evidence
AI Governance Inside Government
Require pre-deployment risk assessments for AI tools procured by state or local agencies
Mandate audit trails for automated or algorithm-assisted government decisions
Require public reporting when AI meaningfully affects eligibility for rights, benefits, or enforcement actions
Prohibit agencies from delegating final decision-making authority to vendors or black-box systems
Establish uniform statewide standards for government AI procurement and oversight
Consumer Protection in Digital Markets
Expand enforcement authority against online fraud and deceptive digital practices
Ban manipulative interface designs like hidden fees and opt-ins that attempt to circumvent informed consent
Increase penalties for digital exploitation targeting seniors, children, or vulnerable consumers
Clarify that consumer protection laws apply fully to digital and AI-driven services
Ban dynamic pricing in stores
Competition, Repairability, and User Choice
Strengthen and enforce right-to-repair laws for digital devices and connected products
Require interoperability standards (e.g., uniform charging ports)
Promote repair access to lower costs and reduce electronic waste