Back to Blog

America’s AI Hiring Laws Are a Well-Intentioned Trainwreck

“Bad decisions made with good intentions are still bad decisions.”
– Jim Collins, How The Mighty Fall: And Why Some Companies Never Give In

TL;DR FAQ: How Are AI Hiring Laws Slowing Down Hiring and Reducing Job Opportunities?

▼ Q: Why are AI hiring laws making employers slower to hire?

A: AI hiring laws add heavy compliance costs, legal risk, and operational delays to routine recruiting tasks. Employers must issue notices, run audits, retain data for years, and prepare for potential litigation, which slows hiring decisions and makes companies more cautious about opening roles at all.

▼ Q: How do AI employment laws impact small and mid-sized businesses?

A: Small and mid-sized businesses are hit the hardest because they lack the legal budgets and compliance teams of large enterprises. Annual AI compliance costs can reach six figures per state, forcing many SMBs to stop hiring in regulated states or avoid using modern recruiting tools altogether.

▼ Q: Are employers legally responsible for AI hiring tools built by vendors?

A: Yes. Under U.S. employment law, employers cannot outsource liability. Even if a third-party vendor built the AI tool, the employer is responsible for how it is used in hiring, including any alleged discrimination. Vendor indemnification rarely provides meaningful protection.

▼ Q: Why do transparency and bias audit requirements backfire in hiring?

A: Public disclosures and bias audits expose proprietary hiring criteria, allow candidates to game systems, and give plaintiffs’ attorneys ready-made litigation targets. Bias audits measure statistical differences, not intent or job-related criteria, often treating correlation as discrimination without context.

▼ Q: Are AI hiring laws causing jobs to move out of certain states?

A: Yes. Companies are increasingly hiring remotely or expanding in states without aggressive AI employment regulations. This geographic arbitrage shifts jobs away from heavily regulated states like California and New York toward states with lower compliance risk and cost.

▼ Q: What actually reduces bias in hiring more effectively than AI regulation?

A: Proven methods include structured interviews, skills-based assessments, clear job-related criteria, and blind resume reviews. These approaches improve fairness without requiring public disclosure of algorithms, annual audits, or expensive compliance regimes.

▼ Q: Where can employers track changing AI hiring laws by state?

A: Orrick’s U.S. AI Law Tracker provides a comprehensive, regularly updated overview of enacted and proposed AI laws across all states. It is one of the most reliable resources for monitoring changes, even if it is not exactly light reading.

Removing bias in hiring is a worthy goal. Creating objective, skills-based processes that focus on qualifications instead of demographics is something most employers and recruiters actually agree with.

But that is not what many of today’s AI employment laws are doing.

Instead, lawmakers with little real exposure to hiring or labor markets have built a regulatory regime that is expensive, fragmented, and often contradictory. The result is fewer job opportunities, slower hiring, and growing pressure on small and mid-sized businesses that cannot afford compliance for compliance’s sake.

Good intentions do not override bad outcomes.

The Compliance Cost Death Spiral

Let’s start with the economics, because that is where these policies quietly fall apart.

Employers operating across states like California, New York, Illinois, New Jersey, and Colorado now face overlapping AI employment laws that apply to the exact same recruiting tools. Resume screening, candidate matching, interview analysis, and even standard applicant tracking systems are pulled into these definitions.

This is not light oversight. It is a six-figure annual compliance burden per jurisdiction.

For a Fortune 500 company, this is annoying overhead. For a company with 100 to 300 employees, it is a slow bleed that never stops. These costs do not create better hires. They do not speed up hiring. They do not measurably reduce discrimination. They simply add friction and risk.

Businesses are left with three rational choices:

  1. Violate the laws unknowingly and hope they are not sued
  2. Stop using AI entirely and fall behind competitors
  3. Stop hiring in regulated jurisdictions

Most companies are quietly choosing the third option.

“It’s Gotta Be a Trainwreck Down There, Right?”

“I mean, it’s gotta be a trainwreck down there, right? I mean, just an absolute casserole of nonsense.”
– Stewie Griffin, Family Guy

That joke lands because it is accurate.

Take California. A single recruiting tool can be classified at the same time as:

  • An automated decision system under employment law
  • Automated decision-making technology under consumer privacy rules
  • A bot that requires disclosure if it communicates with candidates

Each framework brings its own record retention rules, disclosure requirements, audit expectations, and penalty structures. None were designed to work together.

If a company is sued under California employment discrimination law, the exposure is real:

  • Unlimited compensatory damages
  • Punitive damages
  • Mandatory attorney’s fees
  • Class action certification that is relatively easy to obtain

Defending one case can easily cost between $500,000 and $2 million, even if the employer wins.

The Illusion of Vendor Protection

One of the most dangerous assumptions companies make about AI hiring tools is that liability sits with the vendor.

It does not.

Employment law is very clear on this point. Employers cannot outsource liability. If a company uses an AI-enabled system to screen, rank, or evaluate candidates, the employer owns the risk, regardless of who built the technology.

Vendor indemnification clauses rarely offer meaningful protection. They are often capped at contract value and frequently exclude discrimination claims entirely. In a class action, those caps are irrelevant.

The employer is on the hook. Every time.

Transparency That Backfires

Many AI hiring laws focus heavily on transparency, especially public disclosure of how automated systems work. On paper, this sounds reasonable. In practice, it creates a mess.

Mandatory public disclosure:

  • Gives competitors a blueprint to reverse-engineer hiring strategies
  • Teaches candidates exactly how to game screening systems
  • Hands plaintiffs’ attorneys ready-made litigation targets
  • Invites media coverage that treats correlation as proof of discrimination

Forcing companies to publicly disclose proprietary hiring methodology is like forcing Google to publish its search algorithm. No serious business would accept that tradeoff voluntarily.

The Bias Audit Problem Nobody Wants to Admit

Bias audits sound sensible. Test AI before using it. What could go wrong?

A lot.

Most audits compare selection rates across demographic groups. If one group advances at a lower rate, the system is flagged. What audits cannot tell you is why the difference exists.

Applicant pools differ. Experience levels differ. Self-selection differs. Job requirements differ. Bias audits measure correlation, not causation. The law often treats disparity as presumptive discrimination anyway.

That quietly pushes companies away from job-related criteria and toward outcome engineering, which is neither practical nor legal.

The Geographic Arbitrage Effect

Businesses respond to incentives. When one jurisdiction becomes dramatically more expensive without offering any real benefit, companies move.

That is already happening.

Employers are hiring remotely everywhere except the most heavily regulated states. Offices are expanding in Texas, Florida, Tennessee, and Arizona. Headquarters decisions are being reconsidered.

The irony is obvious. Laws designed to protect workers are pushing jobs away from the places they were meant to help.

What Actually Reduces Bias in Hiring

This part should not be controversial.

We already know what works:

  • Structured interviews with consistent scoring
  • Skills-based assessments tied directly to job performance
  • Clear, job-related requirements
  • Blind resume reviews where feasible

None of these require public disclosure of algorithms. None require annual third-party audits. None require six-figure compliance budgets.

Yet current AI laws push employers back toward referrals, brand signaling, and network-based hiring, the exact mechanisms that historically increase bias.

Trying to Keep Up? Enjoy the Light Reading

If you want to track how these laws are changing across states, Orrick maintains a comprehensive U.S. AI law tracker covering enacted laws, proposed bills, and regulatory activity.

Enjoy the light reading: https://ai-law-center.orrick.com/us-ai-law-tracker-see-all-states/

You will need a strong coffee.

A Necessary Disclaimer

This article reflects operational and business perspectives on AI hiring and compliance. It is not legal advice. Employment law is complex, highly fact-specific, and evolving quickly. Companies should always consult qualified legal counsel before making decisions related to AI, hiring practices, or regulatory compliance.

Sources

AI and employment law tracking

Employment law firms

Federal guidance

  • EEOC guidance on AI and employment decision-making

Standards and frameworks

  • NIST AI Risk Management Framework

Industry analysis

  • SHRM compliance briefings on AI in recruiting
  • Brookings and AEI research on automated hiring

Recruiting redefined; built for high-tech,
high-growth teams

Hire Talent
Job Board