AI in the Workplace: What California Small Business Owners Need to Know
- hschwanhauser .
- 2 days ago
- 14 min read
Artificial intelligence is no longer something California employers need to prepare for in the distant future. It is already being used to write emails, summarize documents, screen job applicants, answer employee questions, analyze information, create marketing content, improve customer service, assist with recruiting and perform administrative work.
For small businesses, the accessibility of generative AI has significantly changed the technology landscape.
Companies no longer need large technology departments or specialized programmers to use artificial intelligence. Many AI systems can now be accessed through ordinary software, websites and applications using natural-language instructions.
The result is rapid adoption.
Gallup reported in July 2026 that 52% of U.S. employees use AI in their jobs, with 30% using it at least a few times each week and 15% using it daily. Gallup also found that 47% of employees said their organization had integrated AI tools to improve productivity, efficiency or quality.
Small businesses are participating in the same transition. A 2025 U.S. Chamber of Commerce report found that 58% of small businesses were using AI, up from 40% in 2024 and 23% in 2023. The Chamber reported particularly strong adoption of generative AI chatbots and other tools used to interact with customers and manage business operations.
For California small-business owners, however, using artificial intelligence is not simply a technology decision.
When AI in the workplace is used to recruit employees, screen resumes, evaluate applicants, monitor workers, analyze performance or recommend employment decisions, existing employment and civil-rights laws still apply.
California has also adopted specific regulations addressing automated employment systems.
The practical question for employers is therefore no longer simply:
"Should our company use AI?"
The more important questions are:
Where should AI be used? What information can employees enter into it? How should AI-generated information be verified? When must a human make the final decision? And how can a business prevent technology from creating discrimination, privacy or employee-relations problems?
AI in the Workplace Is Already Changing How Small Businesses Operate
Artificial intelligence refers broadly to computer systems designed to perform tasks that traditionally require aspects of human intelligence.
Modern AI encompasses several different technologies.
Machine learning uses data and algorithms to identify patterns and generate predictions or classifications.
Natural language processing allows computer systems to understand, generate and work with human language.
Generative AI creates new content such as text, images, summaries, computer code and other materials based on user instructions.
Predictive systems analyze data to estimate future outcomes.
Automated decision systems can analyze information and generate scores, classifications, recommendations or decisions.
These technologies are increasingly integrated into ordinary business software.
An employer may therefore be using AI without purchasing something specifically labeled an "AI system."
Recruiting platforms, payroll systems, applicant-tracking software, scheduling programs, productivity applications, customer-service software and performance-management systems may contain artificial-intelligence components.
Where Small Businesses Are Using Artificial Intelligence
The most common business uses of AI generally involve improving or accelerating existing work rather than completely replacing employees.
Examples include:
writing and editing business communications;
creating marketing content;
summarizing reports;
conducting preliminary research;
drafting job descriptions;
reviewing resumes;
scheduling interviews;
communicating with job applicants;
answering routine employee questions;
developing training materials;
analyzing spreadsheets;
preparing meeting summaries;
producing customer-service responses;
translating documents;
identifying patterns in business data;
assisting with inventory management;
generating sales materials; and
automating repetitive administrative tasks.
SHRM reported in 2025 that AI adoption across HR functions was increasing rapidly. Among organizations using AI for recruiting, reported uses included writing job descriptions, screening resumes, automating candidate searches, customizing job advertisements and communicating with applicants. SHRM reported that efficiency was the most commonly identified benefit.
Harvard Business Review has similarly documented the expansion of generative AI into everyday work. Its research and reporting describe AI as increasingly accessible to ordinary employees rather than solely to technology specialists. HBR noted that generative AI can augment, automate or redesign significant portions of U.S. work activity.
The growing consensus is therefore not that every job will be completely automated. Current workplace data show employees using AI primarily to complete specific tasks faster or more effectively.
Gallup's 2026 research found that employees were frequently using AI for activities including writing, research, problem-solving, consolidating information and generating ideas.
AI Does Not Eliminate the Employer's Responsibility
One of the most important principles for California employers is simple:
Using software to make an employment decision does not eliminate the employer's responsibility for that decision.
An employer generally cannot defend a discriminatory hiring or employment decision merely by explaining that software recommended it.
This principle has become increasingly significant as employers use algorithms to screen applicants and evaluate workers.
The U.S. Equal Employment Opportunity Commission has specifically identified technology-related employment discrimination as an enforcement priority. The agency states that federal employment-discrimination laws continue to apply when employers use artificial intelligence, machine learning, algorithms or other automated technologies in employment decisions.
California has gone further.
California's Rules for AI and Automated Employment Decisions
California's Civil Rights Council adopted regulations specifically addressing automated decision systems used in employment.
The regulations became effective October 1, 2025.
They clarify that California's existing employment discrimination protections apply when employers use artificial intelligence and automated systems to make or assist with employment decisions.
According to the California Civil Rights Department, an automated decision system may violate California law if its use harms applicants or employees because of protected characteristics such as race, sex, disability or other protected status.
The regulations also require covered employment records involving automated-decision systems to generally be maintained for at least four years.
California's regulations further address automated tests or assessments that could improperly obtain information about a disability.
For California employers, this means that AI used for hiring, promotion, performance evaluation or other employment decisions should be treated as an HR compliance issue—not merely a software purchase.
AI Hiring Tools Create Particular Risk
Recruiting is one of the areas where artificial intelligence has become especially common.
AI-enabled hiring systems may:
search resumes;
rank applicants;
compare resumes against job requirements;
recommend candidates;
evaluate online assessments;
conduct or analyze video interviews;
identify potential applicants; or
automatically reject candidates based on configured criteria.
These tools can save significant time.
They can also create risks when employers do not understand how the technology reaches its conclusions.
A system trained on historical hiring information can potentially reproduce patterns contained in that information.
For example, if an employer historically hired primarily from one demographic group, an algorithm trained to identify candidates resembling previously successful hires can potentially recreate those patterns even when protected characteristics are not explicitly programmed into the system.
This problem is commonly referred to as algorithmic bias.
Federal regulators have repeatedly warned that employers remain responsible for ensuring that selection procedures do not unlawfully discriminate.
The iTutorGroup Case: Automated Rejection Based on Age
One of the clearest examples of the risks associated with automated hiring systems involved iTutorGroup.
The U.S. Equal Employment Opportunity Commission filed an age-discrimination lawsuit alleging that the company's hiring software was programmed to automatically reject certain older applicants.
According to the EEOC, the software rejected female applicants who were 55 or older and male applicants who were 60 or older.
The EEOC stated that more than 200 qualified U.S. applicants were rejected because of their age.
In 2023, iTutorGroup agreed to pay $365,000 and provide other relief to settle the lawsuit.
The case is particularly significant for small employers because the discrimination was not alleged to have resulted from an individual hiring manager making a biased comment during an interview.
The problem involved the automated screening process itself.
An automated system can therefore create discrimination at scale if inappropriate criteria are incorporated into the technology.
Workday Litigation: Who Is Responsible When an Algorithm Screens Applicants?
A major California case involving Workday illustrates another developing issue.
In Mobley v. Workday, job applicants have alleged that Workday's automated applicant-screening technology discriminated based on characteristics including age, race and disability.
The litigation is ongoing, and the allegations should not be treated as established findings of liability.
However, a federal court in California concluded that the plaintiff had plausibly alleged that Workday could potentially be treated as an agent of employers when customers delegated traditional hiring functions to Workday's algorithmic tools.
The court explained that the allegations involved technology doing more than simply storing applications: the software allegedly participated in recommending candidates and rejecting others.
The case matters for employers because it raises a critical compliance question:
If your company purchases software that screens applicants, who has verified that the screening process is lawful?
A vendor's software contract should not be treated as a substitute for employer oversight.
CVS and AI-Assisted Video Interviews
CVS Health also faced litigation connected to artificial intelligence used during hiring.
A job applicant alleged that CVS required candidates to use HireVue video-interview technology that incorporated AI-based analysis without providing appropriate notice under Massachusetts law.
The applicant alleged that the technology analyzed characteristics including facial expressions and voice-related information and that applicants were not provided appropriate information about the assessment.
CVS reached a settlement in the litigation in 2024.
Although that lawsuit arose under Massachusetts law rather than California law, it provides an important example for California employers: hiring technology can implicate laws other than traditional discrimination statutes.
Privacy, biometric information, consumer-reporting requirements and notice obligations may also become relevant depending on how a system operates.
A New Generation of AI Hiring Litigation Is Emerging
The legal landscape continues to develop.
In 2026, plaintiffs filed litigation against AI hiring technology provider Eightfold AI alleging that its systems collected and analyzed information about job applicants and generated candidate rankings without meeting certain disclosure and consumer-reporting requirements.
The complaint alleges that employers including large national companies relied on AI-generated applicant evaluations. Because these are current allegations, they have not been established as findings of wrongdoing.
The significance for California businesses is that AI employment litigation is expanding beyond traditional discrimination theories.
Courts are increasingly being asked to examine:
algorithmic discrimination;
applicant privacy;
biometric data;
consumer-reporting laws;
notice requirements;
automated ranking;
data accuracy; and
responsibility between employers and AI vendors.
California Privacy Law and Automated Decision making
Employers should also be aware of California's privacy rules governing automated decisionmaking technology.
The California Privacy Protection Agency finalized regulations in 2025 addressing automated decisionmaking technology, commonly called ADMT.
The regulations took effect January 1, 2026, although businesses subject to the ADMT-specific requirements have until January 1, 2027 to comply with those provisions.
For covered businesses, the regulations provide rights relating to significant decisions made using automated technology. Depending on the circumstances, those rights can include notice, access to meaningful information about automated decisionmaking and opportunities to opt out in qualifying situations.
These requirements do not automatically apply to every small California business. The California Consumer Privacy Act has coverage thresholds and exemptions that must be evaluated separately.
However, the development demonstrates California's broader direction toward increased transparency and accountability surrounding automated decisions.
The Risks of Employees Using Public AI Tools
Not all AI risk comes from HR software.
For many small businesses, the more immediate issue is employees independently using generative AI.
An employee may paste information into an AI platform and ask:
"Summarize this."
That information could include:
confidential employee information;
payroll records;
customer information;
medical information;
personnel records;
trade secrets;
attorney communications;
financial information;
proprietary business plans; or
personally identifiable information.
Once information leaves the company's controlled systems, questions can arise about how the AI provider stores, processes or uses that information.
Businesses should therefore establish rules concerning what information employees may enter into external AI platforms.
AI-Generated Information Can Be Incorrect
Generative AI systems can produce information that appears authoritative but is incorrect, outdated or unsupported.
This creates particular concerns in HR.
An employee should not rely exclusively on a chatbot to answer questions such as:
"How much overtime do we owe this employee?"
"Can this employee be terminated?"
"Does this person qualify for protected leave?"
"What is California's minimum wage?"
"Can we classify this worker as an independent contractor?"
Those questions involve laws and factual circumstances that require verification.
AI can assist with research or organization, but important HR decisions should be validated using authoritative sources and qualified HR or legal professionals where appropriate.
Human Review Remains Critical
One of the clearest themes across current research, regulatory guidance and emerging litigation is the importance of meaningful human oversight.
AI can identify patterns.
It can summarize information.
It can draft documents.
It can recommend candidates.
It can organize large amounts of data.
But employment decisions often require context that an automated system may not understand.
For example, an algorithm may identify an employee as having excessive absences.
A human review may reveal that several absences involved protected medical leave.
Software might rank an applicant lower because of gaps in employment.
A human reviewer may determine that the gap involved military service, caregiving or disability-related circumstances.
An automated performance system might identify declining productivity.
A manager may know that the employee recently received an approved accommodation.
Human review helps identify the legal and factual context surrounding the data.
AI Should Not Be Used as an Automatic Termination System
Small businesses should be especially cautious about allowing software to automatically determine discipline or termination.
Even when technology identifies a legitimate workplace problem, the employment decision should generally be reviewed by someone who understands the employee's history.
Before relying on an AI-generated recommendation, an employer should consider:
whether the underlying data are correct;
whether protected leave is involved;
whether the employee has requested an accommodation;
whether the employee recently complained about harassment or discrimination;
whether the employee raised a wage complaint;
whether workers' compensation is involved;
whether policies have been applied consistently;
whether the employee had notice of expectations; and
whether the recommended action is consistent with previous decisions.
Automation may identify a problem.
It should not eliminate factual review.
AI Can Be Valuable in HR When Used Appropriately
The existence of compliance risks does not mean artificial intelligence has no legitimate place in human resources.
AI can be particularly useful for administrative or preparatory work.
For example, an HR professional might use AI to create an initial draft of:
a manager training outline;
an onboarding checklist;
interview questions;
an employee communication;
a performance-review template;
a job description;
a training agenda;
a policy comparison;
a meeting summary; or
an employee survey.
The output can then be reviewed by a qualified person before it is used.
The difference is significant.
AI assists the professional rather than replacing professional judgment.
AI Training Is Becoming an HR Responsibility
Giving employees access to an AI application without training them on appropriate use can create avoidable problems.
Training should explain what AI tools employees are permitted to use and how they may use them.
Employees should understand:
what company information is confidential;
what information cannot be entered into public AI systems;
when AI-generated information must be verified;
how copyright and intellectual-property issues should be handled;
when human approval is required;
whether AI-generated work must be disclosed;
which employment decisions cannot be delegated to AI; and
who employees should contact with questions.
Gallup's research demonstrates why this matters.
In 2025, Gallup found that many employees were using AI without clear organizational guidance. Only 30% said their organization had general guidelines or formal AI policies, even as workplace adoption increased.
Gallup's subsequent research found a strong relationship between manager support and effective adoption. Employees whose managers actively supported AI use were significantly more likely to use the technology frequently and to report that the tools were useful.
The evidence therefore supports both training and governance, rather than simply giving employees access to software.
California Small Businesses Should Consider an AI Workplace Policy
As AI becomes more common, employers may benefit from establishing a written workplace AI policy.
The policy does not need to prohibit artificial intelligence.
Instead, it should establish boundaries.
A useful policy may address:
Approved AI tools
Identify whether employees may use public AI applications or only company-approved systems.
Confidential information
Explain what company, customer and employee information may not be entered into AI systems.
Verification
Require employees to verify AI-generated information before relying on it for important business decisions.
Employment decisions
Explain that hiring, discipline, promotion, compensation and termination decisions require appropriate human review.
Intellectual property
Address the use of copyrighted, proprietary and confidential material.
Accuracy
Require employees to review generated content for factual accuracy.
Security
Identify company requirements for accounts, access controls and data handling.
Disclosure
Determine when employees must disclose that AI was used to create work.
HR oversight
Require HR review before introducing automated technology that will evaluate job applicants or employees.
Questions to Ask Before Buying AI-Powered HR Software
Small employers should perform due diligence before purchasing AI-enabled recruiting or employee-management software.
Questions should include:
What employment decisions does the system influence?
What information does it collect?
How are candidates or employees scored?
What data were used to develop or train the system?
Has the system been evaluated for discriminatory impact?
Can the employer review how a recommendation was generated?
Can a human override the recommendation?
What applicant or employee notices are provided?
How long is employment-related data retained?
Where is the information stored?
Does the vendor use customer information to train its systems?
What happens if an applicant requests an accommodation?
Does the system analyze facial expressions, speech, biometrics or disability-related characteristics?
What documentation will the vendor provide if the employer's practices are challenged?
How will the system comply with California's automated-decision regulations?
The employer should understand the answers before the system begins making recommendations about real employees or applicants.
What the National Research Says About AI and Small Business
Despite differences among studies, several themes consistently appear across major workplace research.
First, AI adoption is growing quickly.
Gallup reported that 52% of U.S. workers were using AI at work by mid-2026.
The U.S. Chamber of Commerce reported that 58% of small businesses were using AI in 2025.
Second, AI is most commonly used to augment tasks rather than completely replace jobs.
Gallup reports frequent uses involving writing, research, information consolidation, idea generation and problem-solving.
Third, manager support and organizational guidance influence successful adoption.
Gallup found substantially greater AI adoption among employees whose managers actively supported the technology.
Fourth, HR use is expanding rapidly.
SHRM reports widespread use of AI in recruiting, including job-description writing, resume screening, candidate searches and applicant communication.
Finally, existing employment laws continue to apply regardless of the technology used.
The EEOC and California Civil Rights Department have both made that principle clear.
AI Is a Business Tool, Not an HR Decision Maker
For small employers, one of the most useful ways to think about artificial intelligence is as another business tool.
Spreadsheets changed accounting.
Email changed communication.
Applicant-tracking systems changed recruiting.
Cloud software changed recordkeeping.
Artificial intelligence is now changing how information is created, analyzed and used.
But technology does not change the employer's underlying responsibilities.
If a hiring decision discriminates unlawfully, the fact that an algorithm recommended the decision does not automatically protect the employer.
If confidential employee information is improperly disclosed, the fact that an employee entered it into an AI chatbot does not make the disclosure harmless.
If a manager disciplines someone based on inaccurate AI-generated information, the employee still experiences the consequences.
The use of AI therefore requires governance as well as innovation.
Preparing Your California Business for AI in the Workplace
California small businesses do not need to become artificial-intelligence experts.
They do need to know where the technology is being used.
A practical starting point is to identify every AI tool currently used by employees and managers.
Then determine:
what the tool does;
what information is entered;
who has access;
whether confidential information is involved;
whether it influences employment decisions;
whether employees have received training;
whether a human reviews important results; and
whether the company's existing policies address AI use.
Businesses should pay particular attention to recruiting and employee-management technology because California's employment regulations specifically address automated decision systems.
An AI policy can then establish expectations for ordinary workplace use.
How Moving Mountains HR Can Help California Employers With AI in the Workplace
Artificial intelligence is creating opportunities for smaller employers to perform work more efficiently, but it is also introducing new HR questions.
California employers increasingly need to determine:
Can employees use generative AI for their work?
What information can they enter into AI systems?
Can AI screen resumes?
Can managers use AI to draft performance reviews?
Who should verify AI-generated employment information?
Does our recruiting software use automated decisionmaking?
Do we need an AI workplace policy?
Are our managers trained on appropriate AI use?
Moving Mountains HR can help California businesses integrate artificial intelligence into workplace processes while maintaining appropriate HR controls.
Support can include reviewing workplace AI practices, developing AI-use policies, training managers and employees, evaluating HR workflows, reviewing recruiting and performance-management processes, updating employee handbooks and helping businesses establish appropriate human review for employment decisions.
The objective is not to remove technology from the workplace.
It is to ensure that technology supports the organization without replacing the judgment, documentation and compliance processes that employers still need.
The Future of AI in the California Workplace
Artificial intelligence will continue to develop, but employers do not need to predict exactly what the technology will look like several years from now.
The current compliance principles are already clear.
AI can improve productivity.
AI can automate repetitive work.
AI can help employees organize and analyze information.
AI can assist recruiting and HR professionals.
But employers remain responsible for employment decisions, discrimination laws still apply, confidential information still must be protected and important employment actions still require appropriate review.
California's new automated-decision regulations make those responsibilities increasingly explicit.
For small businesses, the most effective approach is therefore not unlimited AI adoption and not an outright ban.
It is controlled, documented and informed use of AI with meaningful human oversight.
As artificial intelligence becomes a normal part of work, businesses that establish clear expectations now will be better positioned to use the technology while reducing avoidable employee-relations and compliance risks.
Need Help Creating an AI Workplace Policy?
If your California business is already using ChatGPT, Microsoft Copilot, Gemini, automated recruiting tools or other artificial-intelligence applications, now is an appropriate time to establish clear workplace expectations.
Moving Mountains HR can help you evaluate how AI is being used within your organization, develop an AI workplace policy, train managers and employees, and review HR processes that involve automated decisionmaking.
Contact Moving Mountains HR to discuss practical HR support for your California business.
This article is provided for general informational and educational purposes only and does not constitute legal advice. Artificial-intelligence technology and related laws are developing rapidly. Employers should review current regulatory requirements and consult qualified legal counsel when legal advice is required.
Comments