In 2024, Colorado dipped its toe into AI regulation with passage of Senate Bill (SB) 24-205. The bill, which takes effect on February 1, 2026, attempts to regulate the use of high-risk AI systems. With passage of SB24-205, Colorado became the first state in the country to pass such an encompassing artificial intelligence law. Since then, concerns about this bill’s impact on investment, employment, business operations, and small businesses have been raised, and a second bill, SB318, was tabled during the 2025 General Session after attempting to address the concerns raised about the first bill. 

This report analyzes how SB24-205 is likely to affect both developers and deployers. (For the most part, it does not analyze potential consumer impact, although surely residents also will bear costs.) 

Colorado is a leader in the technology industry: 10% of Coloradans employed are in the industry and the sector sustains 20% of state GDP. Given the hyper-competitive nature of the industry and the effect good and bad regulation can have on economic growth, our data show the state may be well-served treading carefully into the AI regulation arena.

Key Findings 

• Colorado is the first state to pass such a consequential artificial intelligence law, while other states have struggled to find an overall net benefit in regulating developers and users of AI systems.
•  Over 150 bills concerning AI were introduced during the 118th federal Congress (January 2023 through January 2025), but none were passed.
• Colorado ranks third in the nation for tech sector concentration, highlighting its importance as a hub for innovation and economic growth. Every job created in the tech sector supports an estimated 2.67 additional jobs across other industries, showing the ripple effect and broad economic impact of a strong tech ecosystem.
• In terms of AI-related jobs, Colorado also ranks high in job postings, having the 14th highest AI job postings, behind Ohio and just above Michigan. Colorado is expected to continue to be a leader in the tech and AI sector. CSI estimates AI jobs account for approximately 14% of all 2024 job growth in the state.
• Venture capital investment plays an important role in Colorado’s economy and is expected to continue to grow into 2026 and beyond, with Colorado firms capturing a disproportionate share of venture capital funding. In 2024, Colorado ranked fifth in the nation in terms of venture capital funding.
• Over 40 states have some form of AI-related legislation for specific topics, such as deepfakes, illicit images, or music copyright. Among states, 46% have not attempted any framework legislative changes to AI regulation, while according to the International Association of Privacy Professionals, only four have some form of framework AI-related legislation in statute.
• Businesses will be required, after SB 205’s implementation, to adjust their AI systems, conduct risk assessments, prevent algorithmic discrimination, and retrain employees on how to use the updated systems.
• The impact depends on the sector and whether a business is a developer or deployer. 

If SB24-205 goes into effect, CSI modeled the following economic impacts on developers, the broader technology sector, and AI deployers (users). 

AI Developers/Technology Sector Impact

For developers of AI systems, the economic impact (in 2030) includes: 

• Job losses could reach as high as 30,359 when the effects widen to the broader tech sector.
•  Forgone GDP of $5.5 billion.
• Missing Personal Income growth of $8.9 billion.
• Forgone Disposable Personal Income of $2.1 billion.

AI Deployers Impact

To demonstrate that even a minimal cost burden on key sectors could result in significant negative economic consequences under the proposed legislation, CSI modeled a 1% increase in production cost across six industries likely to be critically impacted – Finance, Housing, Healthcare, Education, Insurance, and Legal Services. REMI modeling results show that in 2030 alone, the Colorado AI bill would cause:

• Roughly 40,000 in job loss across the six key sectors studied.  
•  Over a $4 billion loss in statewide Gross Domestic Product (GDP) (i.e., the total monetary value of all final goods and services produced within an economy during a specific time period).
• Nearly a $7 billion loss in economic output (i.e., closely tied to total business sales across the economy, formally economic output is the total economic activity generated during a specific time period which also includes intermediate goods and non-market activities).
• More than a $4.1 billion loss in disposable income.

To examine just the human resource impacts, for almost all industries, complying with requirements on their use of AI for HR-related business decisions impacts available resources for other activities. From just this reallocation of operational costs, the economic impact includes:

• Missing jobs of 1,480 (all figures are the impact in 2030 alone)
•  Forgone GDP of $171 million.
• Missing Personal Income growth of $181 million.
• Forgone Disposable Personal Income of $155 million.

SB24-205 Explained: What the Law Entails & Who It Affects

SB24-205 was the first consequential artificial intelligence law to be passed in the United States. The goal of the bill was to protect Colorado citizens from algorithmic discrimination by regulating the use of high-risk AI. The bill has several requirements: 

• Businesses must employ a risk management system in their AI systems to recognize and alter discrimination risks. 
•  Businesses must conduct impact assessments annually or within 90 days of any changes made to the system covering AI’s intended use, possible risks, data categories, performance metrics, monitoring procedures, and transparency.
• Consumers must be notified if AI is involved in decision-making, about how the system works, and their right to appeal decisions, correct data, and challenge the system if they believe it has been discriminatory.

The bill bifurcates compliance obligations into two groups: developers and deployers. 

• A developer is a person or entity doing business in Colorado that “develops or intentionally and substantially modifies an artificial intelligence system.”
•  A deployer includes any individual or entity doing business in Colorado that uses a high-risk AI system.

Given these two broad definitions, most employers with a connection to the state will be subject to the requirements of SB 205. In order for an employer to protect itself against claims of algorithmic discrimination, as defined in the bill (i.e., ensure they are qualified for the rebuttable presumption that they used reasonable care), the employer must:

• Implement a reasonable risk management policy regarding their use of AI systems.
•  Complete an annual impact assessment of their AI use or contract with a third-party to complete the assessment.
• Provide notice to employees of their use of high-risk AI systems for making consequential decisions.
• Offer an explanation to each employee of an adverse consequential decision based on their use of AI systems.
• Offer an employee an opportunity to appeal a decision for human review.
• Provide notice to the Attorney General within 90 days if the employer finds that their AI systems have caused algorithmic discrimination.

In addition to the above-mentioned impacts on deployers (users) of AI systems, the bill also requires developers of AI systems to: 

• Provide a statement describing the foreseeable uses of their AI system and known harmful or inappropriate uses of the high-risk AI systems.
•  Documentation that includes:
• Information on the types of data used to train the high-risk AI systems, limitations of the system, and other information necessary for users of the systems to comply with the user requirements.
• How the high-risk AI system was evaluated for performance and how it addresses potential algorithmic discrimination.
• Information that would guide the user in understanding the output of the system.
• Information sufficient for a user to complete an impact assessment.

These requirements, of course, would have economic impacts on individuals and businesses addressed in the following pages. It is worth noting that most of this research focuses on the impact on businesses and users of AI tools. There is also a consumer impact with its potential associated cost-benefit impact that is generally not addressed here.

Understanding What Constitutes a Consequential Decision in SB 205

The background on SB 205 still has open-ended questions on how it would be applied across the multiple dimensions of decision-making. The following Figure 1 provides 31 examples of consequential decisions that could be impacted by SB 205, once enacted. For instance, the use of a high-risk AI system in making a consequential decision regarding employment – such as screening resumes for job interviews – would fall under the regulations of the bill. In contrast, the use of AI for recommending a movie on a streaming platform would not be impacted by the bill. 

The use of an AI system for determining mortgage eligibility, or diagnosing a disease, or tracking computer activity at work would be subject to the bill, while the use of AI for adjusting email marketing timing, sorting through environmental data, or sending housing related messages/reminders would not be regulated by the bill. In general, a consequential decision is a decision that influences the provision or cost of essential services such as education, employment, financial services, healthcare, housing, or legal services – essentially decisions that can materially influence individuals' access to opportunities and resources.

Figure 1

 

The Economic Impact of AI

Given the research presented so far, CSI employed Regional Economic Models Incorporated’s (REMI) dynamic multiplier system to estimate the economywide impacts for developers and deployers, presented in the following two sections.

Regulatory Impact

Before delving into the details of the economic impact, it’s worth asking the question: Do regulations really have any impact on the economy? Perhaps the regulations have no effect? The answer is that regulations typically slow economic growth. For instance, researchers found that compliance and documentations costs in the European Union after the EU’s General Data Protection Regulation (GDPR) took effect saw monthly venture deals drop 17-26% and approximately 34% to 40% less raised per deal compared to the U.S. Essentially, the GDPR regulations slowed the time-to-market for technology products. A separate study found that data access constraints from GDPR led to an exit of mobile app development and halved entry shortly after rollout, consistent with higher fixed costs of compliance and data access restrictions. A third study found that liability and higher-risk obligations can cause a shift away from riskier domains. In essence, consumers get safer systems at the expense of higher costs and slower market introduction of technology systems. At least, that was the finding for medical implants.  Lastly, regulatory fragmentation can lead to market share loss, suggesting a need for harmonization in some standards for optimal growth, at least according to the European Union Parliament regarding their passage of the EU AI Act. 

Broader Economic Impact from the Developer/Tech Sector Impact If the Bill Goes into Effect As Is

Colorado is a leader in the technology sector, with approximately a tenth of its employment base in technology. In terms of AI-related jobs, Colorado also ranks high in job postings, having the 14th highest AI job postings, behind Ohio and just above Michigan (Figure 3). Colorado is expected to continue to be a leader in the tech and AI sector.

Although job postings are typically much higher than actual job growth, the 15,583 AI-related jobs postings that existed in Colorado in 2024 show the AI industry’s potential to be a major driver of employment in the state. In fact, overall job growth in Colorado in 2024 was 22,100. Even if actual AI-related jobs are only one-fifth of the job postings, t number still represents 14% of all job growth in the state.

Figure 2

Regulating AI in the state will have some impact on the aforementioned AI employment base. The magnitude of the impact is dependent on Colorado businesses and employees’ productivity relative to other states and countries.

How, exactly, will SB 205 impact AI developers if it goes into effect? What is the chain of causation in how it could impact developers? Is it possible that AI developers could be unaffected? The answer to these questions is it depends. 

The practical impacts that will cause change for at least some developers include: 

• Higher documentation burden on model inputs, bias mitigation strategies, and fallback procedures.
•  A higher legal and regulatory risk that could result in unfair trade practice accusations.
• Increased operational overhead by requiring developers to stay up to date with updates, issue timely notifications, and maintaining ongoing documentation – all of which can slow development cycles and continuous deployment.
• Although there are exceptions for small businesses, there is the potential for an unequal burden on startups and small developers, specifically in such sectors as human resources and financial resources. 

Suppose for the moment, that the possible impacts are a drop in AI-related employment growth of 5%, 10%, or 15%. How would that impact the broader Colorado economy? 

The following Figure 4 has a comparison. In the scenario where there is a 5% drop in AI-related job growth at a 1:1 ratio of actual jobs to jobs postings, the economywide impact is a drop of 2,734 jobs. The 10% and 15% scenarios are drops in jobs of 5,468 and 8,206. The far-right column has the economywide impact if the actual jobs to jobs postings turns out to be a 5:1 ratio of actual jobs to jobs postings.

 

Figure 3

The just-mentioned table, of course, gives a range of impact. As a more refined estimate, CSI used the minimum impact reported by Jia, Jin, and Wagman (2018) on the impact to the EU from implementation of the GDPR regulation. They found a minimum drop in weekly venture capital deal amount of 17% - meaning a drop in potential growth. 

Venture capital investment plays an important role in Colorado’s economy and is expected to continue to grow into 2026 and beyond, with Colorado firms capturing a disproportionate share of venture capital funding. In 2024, Colorado ranked fifth in the nation in terms of venture capital funding. Investors – be they small or large investors – provide funds for individuals with innovative ideas, supporting high-growth startups, and creating challenging and rewarding jobs. Colorado is a hub for technology innovation, bioscience research, aerospace, and clean energy, and venture capital provides early-stage funding to these industries looking to expand. Startups backed by venture capital often scale faster than traditional businesses, and this scaling creates job opportunities for the ambitious younger generation and throughout the economy’s supply chain. This growth attracts workers skilled enough to compete with workers anywhere in the world, strengthens the state’s entrepreneurial environment, and helps retain talent that could migrate to larger markets like San Francisco, New York City, Phoenix, or Seattle.

In addition to direct job creation, venture capital investment enhances Colorado’s long-term economic well-being by diversifying the economy and fostering well-rounded strength to withstand adverse economic shocks. As companies expand, they increase state tax revenues, attract further investment, and create industry clusters that benefit other businesses along the supply/service chain. When Colorado-based businesses experience successful exits, owners and investors recycle created wealth back into the local economy as founders and employees invest in new innovative ventures of the up-and-coming generation, expanding a positive cycle of innovation and growth. Overall, venture capital not only provides much needed capital for individual companies and their innovators but also strengthens Colorado’s competitiveness, ensuring that the state remains an attractive destination for both entrepreneurs and investors.

Colorado is currently a national leader in the tech industry, with technology jobs accounting for 10% of all statewide employment and contributing 20% to the state’s total GDP. Colorado also ranks third in the nation for tech sector concentration, highlighting its importance as a hub for innovation and economic growth. Every job created in the tech sector supports an estimated 2.67 additional jobs across other industries, showing the ripple effect and broad economic impact of a strong tech ecosystem. This makes the health and stability of Colorado’s tech industry essential – not only for short-term economic performance but also for long-term statewide prosperity.

Given the enormous amount of uncertainty surrounding AI regulation, CSI modeled the potential economic impact for the developer portion in three parts. The first part was to estimate a minimum economic impact, covering only the impact from reduced future growth in venture capital investment. To do this, CSI employed an estimate produced by Jia, Jin, and Wagman (2018) on the impact for venture capital deal volume in the European Union after passage of the European General Data Protection Regulation. The minimum estimated impact was a reduction of 17% in weekly deal volume. CSI used this assumption as a reduction in the growth in intellectual property investments. The estimated growth was from Regional Economic Model’s Incorporated’s (REMI) assumptions on growth in GDP from 2026 to 2030. These two numbers—REMI’s projected GDP growth and a 17% drop in intellectual property investments—was combined with the assumption that Colorado’s venture capital investments are estimated at $10,426 per $1 million in GDP. 

For instance, assuming GDP growth in the state will be $28 billion for 2026. CSI exploits this $28 billion in growth (from REMI) to estimate the projected venture capital investments in the state for 2026 using the assumption that venture capital investment is $10,426 per $1 million in GDP. This assumption produces expected growth in venture capital investment of $292 million. Next, assuming the minimum impact estimated by Jia, Jin, and Wagman (2018) is correct at 17%, then we know the forgone venture capital investment for 2026 is approximately $50 million. This is the minimum impact from just the forgone venture capital investment.

Overall, the results of this analysis show a minimum economic impact from just the developers aspect of the bill of slower economic activity (all figures below are observed for the year 2030):  

• Slower employment growth of 328
•  Missing GDP growth of over $61 million
• Missing business sales (output) of $100 million
• Forgone Personal Income of $43 million
• Forgone Disposable Personal Income of $37 million

A more likely scenario of the impact also stems from Jia, Jin, and Wagman (2018), who also report a 39.6% drop in amount raised on an average deal. Using this figure as opposed to the 17% figure above, the economic impact is as follows: 

• Slower employment growth of 764
•  Missing GDP growth of $143 million
• Missing business sales (output) of $233 million
• Forgone Personal Income of $100 million
• Forgone Disposable Personal Income of $86 million

The impact from the missed opportunity of venture capital investment in Colorado’s startup ecosystem is, of course, just the tip of the iceberg of potential impact. Accounting for 20% of the state’s GDP, throttling the tech sector’s growth will also impact GDP directly. Assuming the 40% trimming on the tech sector’s impact on GDP growth, the direct GDP impact from AI regulation could grow to (all figures in 2030): 

•  Slower employment growth of 30,359
•  Missing GDP growth of $5.5 billion
• Missing business sales (output) of $8.9 billion
• Forgone Personal Income of $4.1 billion
• Forgone Disposable Personal Income of $2.1 billion

Overall, as one would expect, higher documentation and innovation costs slow the attractiveness of the state, slowing investment and overall economic growth (Figure 5).

 

Figure 4

Broader Economic Impact from the Deployer Impact

The second aspect of compliance is the users of AI-related technology for making consequential decisions. This impact varies widely by industry and job tasks. In general, most job-related tasks are unaffected. Outside of a company’s employment practices, most industries have zero to minimal impact from the bill. The industries most heavily affected beyond the HR aspect are listed in Figure 6. The industries with a disproportionate impact include education, employment, financial services, government services, healthcare, housing, insurance, and legal services.

 

Figure 5

SB 205’s Effect on Human Resources for All Industries

An aspect that affects potentially all industries is the use of AI in making consequential employment-related decisions, including: 

• Hiring and firing
•  Performance evaluations
• Promotions
• Background checks
• Electronic monitoring on networks/devices

The likely HR roles that will be impacted include: 

• HR directors
•  Recruiters
• Compliance/legal officers
• Data privacy officers

Broadly, the types of HR-related costs employers may incur include training, policy development, documentation and audits, staffing (compliance officer, legal counsel), and appeals. The following figure provides broadly estimated costs for each of these functions.

 

Figure 6

These broad HR compliance costs capture categories of tasks performed by HR professionals that could be impacted by the use of AI. For a broader look at the HR tasks performed by human resources professionals, Appendix B has details as identified by O*NET database. 

According to Gartner, a research and advisory firm focusing on business and technology topics, companies spend, on average, $2,810 per employee annually on HR functions, or approximately 1.5% of organizational operating expenses or 0.8% of revenue. Breaking down the $2,810, Gartner reports that companies spend $401 per employee on recruiting, $221 on total rewards, and $202 on Learning and Development (L&D). The HR-related line item that is most closely related to the topic of this research – technology – accounts for 8.4% of total HR budget. According to Gartner, this is the top area of planned HR investment for the year.

Given this background, what will be the direct impact across the economy and all employers in the state in complying with the HR aspect of SB 205? 

Although unknown, a reasonable range could be an increase in the HR-related technology spending of 5% to 10%. That equates to a direct, HR-related economywide cost of between $49.7 million and $99.4 million. A recent study from Eloundou, Manning, Mishkin, and Rock (2023) suggests that around 80% of employees could have at least 10% of their tasks affected by AI Large Language Models (LLMs). This 10% is used in the results reported in the following table. Using REMI with a $99.4 million increase in production costs, the economywide economic impact – direct plus indirect and induced – comes to (2026 impact is the year referenced in the following points): 

• Missing employment growth of 1,290
•  Slower GDP growth than otherwise by $79.2 million
• Lower business sales by $137.2 million
• Forgone Personal Income of $109.1 million
• Forgone Disposable Personal Income of $93.2 million

 

Figure 7

Broad Scope: Modeling Sectors with High Impact Based Upon Using AI to Make Consequential Decisions 

If implemented, Colorado’s AI Act is expected to have significant implications for various industrial sectors, particularly those characterized by high exposure to automation and algorithmic decision-making. Findings from a recent working paper by researchers at the University of Pennsylvania suggested that approximately 80% of U.S. workers could see at least 10% of their job tasks affected by AI-driven automation and regulatory requirements. More notably, nearly 20% of workers may experience displacement of over 50% of their job responsibilities, particularly in roles heavily reliant on cognitive and administrative functions.

Further supporting this analysis, complementary research highlights that the majority of these exposed job functions are concentrated among highly educated, high-income, white-collar workers, such as professionals in finance, law, healthcare, and legal services. These occupations not only involve extensive use of decision-support algorithms but also fall within the scope of the "high-risk" AI systems regulated under the Act. As a result, compliance burdens, operational adjustments, and potential reductions in workforce needs could be disproportionately felt within these sectors, amplifying economic disruptions and productivity risks across the state – possibly making Colorado a less attractive place for businesses and professionals to move to.

Sector-Specific Impact Methodology 

To assess the potential economic impact of Colorado’s AI Act going into effect, CSI utilized the REMI modeling framework, focusing on Production Cost as the policy variable of interest. The analysis assumes that regulatory constraints of the AI bill will lead to increased administrative costs across key industries in the state. Production Cost is a critical input in regional economic modeling because it directly influences a firm’s competitiveness, profitability, and operational decisions. In the context of the Colorado AI Act, the regulation of AI technologies is expected to:

• Reduce automation efficiency in routine tasks, increasing reliance on human labor and manual processes;
•  Raise compliance and administrative costs as firms adapt to new reporting and oversight requirements; and
• Slow innovation and adoption of productivity-enhancing AI tools, reducing output per worker.

These factors would increase production costs across multiple industries. By modeling this variable, CSI captured how these higher costs ripple through the economy – impacting pricing, investment decisions, employment, and overall output.

Furthermore, higher costs of production can make Colorado-based firms less competitive relative to those in other states or countries with less restrictive AI environments, potentially leading to relocation, lower demand for businesses and professionals to locate here, reduced investment, and/or contraction in high-value sectors. 

Regulatory compliance typically imposes additional cost burdens on the industries and organizations subject to the legislation. To assess the potential magnitude of the Colorado AI bill's economic impact, CSI modeled a highly conservative 1% increase in production costs across several AI-sensitive industries: Financial Services, Housing, Healthcare, Education, Insurance, and Legal Services. These sectors are among the most likely to face significant economic friction as a result of the new legislation. 

CSI selected a 1% increase in production costs on relevant industries for three key reasons:

• To maintain a conservative approach in estimating potential economic impacts;
•  Due to limited available data and the high level of uncertainty surrounding the bill’s downstream effects on industries and the broader economy; and
• To demonstrate that even a minimal cost burden on key sectors could result in significant negative economic consequences under the proposed legislation.

Lastly, the cumulative impacts mentioned in the following sections refer to the sum of annual numbers from 2026 through 2036. The jobs numbers are essentially job-years and the GDP numbers GDP-years. To get a estimate of the annual impact from the cumulative figures, please refer to the table or if a rough estimate works, divided the cumulative impact by 11.

Financial Services Sector Impact

The Financial Services industry – classified within the broader Business and Financial Services super-sector – represents a critical component of Colorado’s economy. According to data from the 2024 Occupational Employment and Wage Statistics (OEWS), the Business and Financial Services sector employed approximately 270,120 individuals, making it the fourth-largest employment sector in the state. 

Financial Services in particular is likely to be significantly impacted by Colorado’s AI bill due to its reliance on consequential decision-making – defined in this case as decisions that affect an individual's legal rights, financial status, or access to services.

Common activities such as loan approvals, credit scoring, insurance underwriting, and fraud detection all involve consequential decisions. When AI is used to assist in these processes, firms must comply with new requirements, including transparency, impact assessments, and documentation. Given the sector’s dependence on data-driven tools, the bill introduces potential compliance costs, operational challenges, and legal exposure, especially for firms operating across multiple jurisdictions.

In 2024, the average annual wage for workers in this sector was $99,316, which is approximately 54% higher than the estimated 2025 living wage for a single adult in Colorado. This wage premium underscores the sector’s concentration of high-skill, high-income white-collar occupations—roles that research has identified as disproportionately exposed to AI-driven task automation and regulatory scrutiny under the Colorado AI Act., 

Given its scale, wage levels, and vulnerability to AI regulation (particularly in compliance, risk modeling, and customer analytics), the financial services sector is expected to face significant disruption under both moderate and high-impact modeling scenarios. Implications include potential employment contractions, reduced productivity, and elevated compliance costs—especially for small and mid-sized firms without dedicated legal or technical infrastructure to meet new regulatory demands. 

Tasks that would likely most impact financial workers due to Colorado’s unique approach to AI regulation include: 

• Recruiting staff members;
•  Overseeing training programs;
• Approving, rejecting, or coordinating the approval or rejection of lines of credit or commercial real estate, or personal loans; and
• Examining, evaluating, or processing loan applications. 

Figure 8 below reports the REMI results for the 1% production cost increase scenario. Estimates begin in 2026, when the bill is intended to be implemented, and projects out 10 years to 2036. According to CSI’s projections, between 2026 and 2036 alone, increasing operational costs for the Financial Services sector by only 1% due to AI regulatory burdens would cause:

• Nearly 10,000 jobs lost in 2030 alone;
•  Nearly $9.7 billion cumulative loss in statewide GDP;
• Over $16 billion cumulative loss in statewide economic output; and
• More than $10 billion cumulative loss in disposable personal income. 

 

Figure 8

Real Estate, Rental, & Leasing Sector Impact 

When individuals negotiate a rental agreement, apply for a loan, or face eviction – all of which are consequential decisions – the individual making the decision, be it the potential landlord, financial intermediary, or current landlord, may use some type of AI to assist in their decision-making process. This use could be regulated. 

To estimate the bill’s potential impact on the housing industry, CSI modeled economic impacts for the Real Estate, Rental, and Leasing industry. This industry – classified within the broader Sales and Related super-sector – represents 305,140 workers in the state and makes it the second-largest employment sector in Colorado. 

In 2024, the average annual wage for workers in this sector was $64,493, which is approximately 16.3% higher than the estimated 2025 living wage for a single adult in Colorado., 

While this wage level indicates relative income stability for many workers, the sector's exposure to AI regulation arises from the growing use of automated systems in credit scoring, tenant screening, property valuation, and mortgage underwriting – all of which involve consequential decisions that may fall under the Act’s high-risk AI classification.

As a result, real estate, rental, and leasing firms – particularly those relying on AI for lead generation, predictive pricing, or applicant evaluations – may face significant compliance demands, including algorithmic impact assessments, increased transparency obligations, and enhanced consumer appeal processes. For smaller agencies and property managers, these requirements could translate into notable cost burdens, increased legal exposure, and a possible reduction in automation-driven efficiencies, ultimately affecting both employment levels and housing market fluidity.

Generally, the largest impact on the real estate sector and its workers includes: 

• Finding potential buyers; and 
•  Portfolio and neighborhood management. 

Figure 9 below reports the REMI results for the 1% production cost increase scenario. Estimates begin in 2026, when the bill is intended to be implemented, and projects out 10 years to 2036. According to CSI’s projections, between 2026 and 2036 alone, increasing operational costs for the Real Estate, Rental, and Leasing sector by only 1% due to AI regulatory burdens would cause:

• Over 11,000 jobs lost in 2030 alone;
•  Nearly $5.4 billion cumulative loss in statewide GDP;
• Over $11 billion cumulative loss in statewide economic output; and
• Nearly $11 billion cumulative loss in disposable personal income. 

 

Figure 9

Healthcare Sector Impact 

The Healthcare industry is another key sector likely to experience substantial regulatory impacts under Colorado’s AI Act, due primarily to the high frequency of consequential decision-making inherent in clinical and administrative functions. These decisions – ranging from patient diagnostics to treatment eligibility and scheduling – fall squarely within the Act’s definition of high-risk AI applications, making the sector particularly sensitive to both compliance burdens and operational shifts.

In Colorado, healthcare employment spans two major occupational categories: Healthcare Practitioners and Technical Occupations and Healthcare Support Occupations. Combined, these sectors employed an estimated 259,360 workers as of 2024. According to OEWS data, the average annual wage across these roles is approximately $76,417, which is about 30% higher than the projected 2025 living wage for a single adult in the state.,

Tasks that would most likely impact healthcare workers include: 

• Selection of new patients via IT services; 
•  Analyzing patient's medical history, medication allergies, physical condition, and examination results to verify operation's necessity and to determine best procedure;
• Informing parents and guardians of health problems and surgical procedures through various channels, such as in-person and telecommunication systems;
• Referring patients to medical specialists or other practitioners when necessary, assuming IT systems are used;
• Interpretation of results from tests and physical examinations if IT systems are utilized; and
• Preparing case histories if IT systems are utilized. 

Figure 10 below reports the REMI results for the 1% production cost increase scenario. Estimates begin in 2026, when the bill is intended to be implemented, and projects out 10 years to 2036. According to CSI’s projections, between 2026 and 2036 alone, increasing operational costs for the Healthcare sector by only 1% due to AI regulatory burdens would cause:

• Over 13,300 jobs lost in 2030 alone;
•  Nearly $18 billion cumulative loss in statewide GDP;
• Over $29 billion cumulative loss in statewide economic output; and
• Nearly $15 billion cumulative loss in disposable personal income. 

 

Figure 10

Among the sectors analyzed, overall, the Healthcare industry is most at risk of higher regulatory costs from the current AI statute. This makes sense given the industry’s size, the broad range of services it offers, and the high number of consequential decisions subject to the new AI regulations. 

These results suggest that the Colorado AI bill is likely to cause hospitals and healthcare systems to divert resources away from direct patient care. Consequently, financial and human resources may be reallocated from clinical functions to legal, IT, and compliance departments. Furthermore, concerns over legal liability and unclear regulatory guidelines could lead healthcare organizations to delay or reduce adoption of AI tools that currently enhance efficiency and patient outcomes. 

These disruptions may slow innovation, exacerbate existing operational inefficiencies, and strain already limited healthcare resources. Collectively, these factors may reduce the system’s capacity to deliver timely, high-quality care, even as demand for healthcare services continues to rise.

Education & Vocational Training Sector Impact 

The Education and Vocational Training sector in Colorado is also a sector expected to face significant disruption as a result of AI regulation. As with other industries examined in this report, the disruption is primarily driven by the sector's reliance on frequent, high-stake decision-making. For the purposes of modeling these impacts, CSI has grouped K–12 and Higher Education into a single category.

In Colorado, the Education and Vocational industry falls into the larger Educational Instruction and Library sector and employs an estimated 154,800 Coloradans. These workers are associated with an average annual wage of $66,892 – roughly 20% higher than the estimated living wage in the state for a single adult in 2025.,

Tasks that would likely impact education sector workers include: 

• Utilizing IT grading software; 
•  Developing teaching plans with software programs;
• Adapting teaching methods and instructional materials to meet students' varying needs and interests;
• Using computers, audio-visual aids, and other equipment and materials to supplement presentations; and
• Administering standardized ability and achievement tests through IT software which interpret results to determine student strengths and needs. 

Figure 11 below reports the REMI results for the 1% production cost increase scenario. Estimates begin in 2026, when the bill is intended to be implemented, and projects out 10 years to 2036. According to CSI’s projections, between 2026 and 2036 alone, increasing operational costs for the Education and Vocational sector by only 1% due to AI regulatory burdens would cause:

• Over 1,440 jobs lost in 2030 alone;
•  Over $1.5 billion cumulative loss in statewide GDP;
• Over $2.5 billion cumulative loss in statewide economic output; and
• More than $1.4 billion cumulative loss in disposable personal income. 

 

Figure 11

These results indicate that the Colorado AI bill could increase regulatory and operational costs for educational institutions, prompting a shift in resource allocation away from instructional spending – particularly teacher salaries – toward compliance-related expenditures. 

As schools and universities adapt to new oversight requirements for AI systems used in curriculum planning, grading, admissions, and student services, they may be forced to invest in legal support, administrative oversight, and upgraded technology systems to ensure compliance. These additional cost pressures can divert funding from core instructional budgets, particularly in public institutions already operating under tight fiscal constraints. As a result, schools may freeze or reduce teacher compensation, delay hiring, or increase class sizes to absorb the financial burden, ultimately weakening educational outcomes and workforce retention in the sector

Insurance & Related Carriers Impact

The Insurance and Related Carrier sector, categorized under Business and Financial Operations Occupations, plays a critical role across nearly every sector of Colorado’s economy. These positions typically require a highly educated workforce and offer competitive salaries to match. For instance, Insurance Underwriters currently earn an average annual wage of $102,991 – approximately 47.6% above Colorado’s 2025 livable wage for a single adult.,

Key functions within the industry – such as the issuance of policies, adjudication of claims, and determination of premiums – are likely to be classified as consequential decisions under the new AI bill when artificial intelligence is used to assist human decision-makers. As a result, the sector may face increased compliance costs and operational disruptions, including potentially reducing human capital to comply with higher expenditures. 

Tasks that would most likely be impacted include: 

• Process and record new insurance policies and claims; 
•  Reviewing insurance policy to determine coverage if IT software is involved;
• Transmitting claims for payment or further investigation if IT software is involved;
• Reviewing and verify data, such as age, name, address, and principal sum and value of property, on insurance applications and policies if IT software is involved;
• Transcribe data to worksheets and enter data into computer for use in preparing documents and adjusting accounts;
• Process, prepare, and submit business or government forms, such as submitting applications for coverage to insurance carriers; and
• Enter insurance- and claims-related information into database systems. 

The insurance sector is the only industry in this analysis to show positive GDP and economic output effects in the first year of implementation, 2026. This initial growth is likely driven by a surge in demand for new or expanded insurance coverage as companies seek to comply with SB 205’s regulatory requirements. However, beyond 2026, the sector experiences negative economic impacts, as the broader ripple effects of the legislation, such as higher compliance costs and reduced efficiency, begin to outweigh early gains.

Figure 12 below reports the REMI results for the 1% production cost increase scenario. Estimates begin in 2026, when the bill is intended to be implemented, and projects out 10 years to 2036. According to CSI’s projections, between 2026 and 2036 alone, increasing operational costs for Insurance and Related Carriers by only 1% due to AI regulatory burdens would cause:

• Nearly 2,600 jobs lost in 2030 alone;
•  Over $2 billion cumulative loss in statewide GDP;
• Nearly $2.9 billion cumulative loss in statewide economic output; and
• Approximately $3 billion cumulative loss in disposable personal income. 

 

Figure 12

Legal Services Sector Impact

Legal services could also be impacted in that the advice of an attorney would likely be considered a consequential decision if the advising individual used AI when giving advice that affects an individual’s rights, freedoms, or legal status. 

This industry is another that is associated with a highly skilled, highly educated, and highly demanded workforce and that touches all corners of any economy. Legal services are a part of the justice, public order, and safety larger sector which has an average annual wage of $94,744. That number exceeds the Colorado 2025 livable wage rate by 43% for a single individual living in the state. , 

If lawyers or legal staff use AI tools their firms may be subject to strict compliance requirements, including transparency in how AI informs decisions, bias assessments, and accountability protocols. Given the sensitive nature of legal outcomes and the high risk of unequal treatment or due process concerns, the bill could create operational slowdowns, increase liability risks, and raise compliance costs, especially for firms adopting AI to manage high caseloads or reduce costs.

Tasks that would most likely impact legal sector workers include: 

• Case evaluations and legal strategy;
•  Plea bargains or settlement advisories;
• Document review in litigation or discovery; and
• Eligibility assessments for legal aid or defense. 

Figure 13 below reports the REMI results for the 1% production cost increase scenario. Estimates begin in 2026, when the bill is intended to be implemented, and projects out 10 years to 2036. According to CSI’s projections, between 2026 and 2036 alone, increasing operational costs for Legal Services by only 1% due to AI regulatory burdens would cause:

• Over 1,000 jobs lost in 2030 alone;
•  Nearly $412 million cumulative loss in statewide GDP;
• Roughly $821 million cumulative loss in statewide economic output; and
• Almost $1.2 billion cumulative loss in disposable personal income. 

 

Figure 13

How Other Studies have Attempted to Model the Economic Impact of AI Regulation at the State Level

CSI reviewed studies on the economic impact of AI regulation at the region and state level. The following are selected results from reviewed studies/bills.

• Florida: Although based solely on an empirical assumption of a 1% drop in productivity, the James Madison Institute found that AI regulation may reduce Florida’s economy by $38 billion. The 1% assumption is similar to estimates published by the IMF, Goldman Sachs, and McKinsey who put the productivity growth from AI adoption of between +0.1% and +1.5%. Essentially, the 1% assumption is assumed missed productivity growth rather than a decline in productivity. If this were to occur across Colorado, it would equate to the following impacts as reported in Figure 15. 
•  Figure 15’s positive impacts appearing in the early years of the simulation are due to a core assumption in the REMI model: when productivity declines, firms must hire more workers to maintain output levels. This initially leads to short-term job growth. However, as time progresses, missing productivity gains means the state becomes less productive overall, lowering efficiency and competitiveness.  As a result, businesses gradually shift operations and investment to more productive states, causing long-term declines across all columns showing economic impacts – including employment, GDP, and economic output – within Colorado.

 

Figure 14

• Florida Senate (SB 936): The bill requires a study every three years on the economic impact of automation, AI, and robotics on employment in the state. This Florida fiscal impact assessment sources an MIT study that suggests that for “each additional robot per thousand workers reduces the local employment-to-population ratio by 0.39 percentage points and wages by about 0.77 percent”. The fiscal assessment also reports on one study that suggests the Miami-Dade, Broward, and Palm Beach counties may see a 23% reduction in their total workforce displaced by automation by 2030. Although the studies will turn out to be inaccurate, they suggest that the impact of algorithmic learning may have a wider impact than what society saw with the onset of the Industrial Revolution.
•  Texas House (HB 2060): This bill established a dedicated council meant to oversee AI implementation within Texas state agencies. It has established a structured framework to evaluate and guide the usage of AI in the state government as the state commits to responsible and ethical AI regulation. The Texas AI Advisory Council was established through this bill to carry out the functions outlined in the bill. The council has not yet released any research or partaken in any studies that are accessible to the public currently.

Internationally, job listings including at least one AI-related skill have been increasing. Australia, the United States, the United Kingdom, New Zealand, and Canada were the first five countries to begin recording job listings that included these skills in 2014, and now 16 more countries have joined this list. Figure 16 visualizes this increase in AI usage in the job market. As AI becomes more prevalent, employers across the world have begun to demand more AI-related skills, indicating AI’s growing influence on the job market and the importance of being able to compete for AI-related business.

 

Figure 15

Other States’ Approach to AI

Colorado is first among states to bring this type of AI regulation forward. Our understanding is that this initial piece of legislation (SB 205) was intended to be a starting point, with future legislation intended to clarify Colorado’s approach within the broader regulatory environments across other states. If Colorado’s approach when implemented is significantly different than other states and countries (such as the EU), companies selling products would either have to have different versions for their product or not sell their product in the state. Colorado-based AI developers would also need to potentially have multiple products. To date, Colorado has led the charge for AI regulation, with California, Utah, and Texas following suit. Figure 17 displays the status of AI-related legislation by state according to the International Association of Privacy Professionals (IAPP). As shown, only four states have some type of legislatively adopted regulation, with most attempts at regulation failing. Stanford University’s Artificial Intelligence Index Report offers a different accounting of AI legislation, suggesting 15 states have adopted some form of AI legislation from 2016 through 2023. The differences stem from how the research groups classify AI legislation. Although different in accounting, the general conclusion is that a minority of states have nothing on the books when it comes to AI legislation.

Of the states that have some type of passed AI regulation framework, California’s has been the most similar to Colorado’s in that it requires AI transparency from developers to clearly disclose and, at no cost, give options to customers when AI is being used. Utah’s bill requires AI disclosure, and Texas assesses potential discrimination in state agencies' AI systems. Over 40 states have some type of AI legislation on the books, but the legislation is in specific areas, such as deepfakes, illicit images, or music copyrights. Among all states, 46% of states haven’t introduced any AI-related framework bills according to IAPP, and 28% of states currently have inactive AI-related framework bills (see Figure 17). With that said, over 40 states have addressed specific AI-related topics, such as deepfakes, illicit images, or music copyrights.

 

Figure 16

Competition in AI

The field of AI is still new. It is complex and hard for a state to do alone without alienating an industry and causing it to be driven elsewhere. In a state known for fierce competition and rapid innovation, poorly designed or overly burdensome regulations risk driving away investment and stifling growth. When compliance becomes too costly or unclear, startups and small businesses are especially vulnerable, reducing the overall competitiveness of the ecosystem. This also highlights a broader issue: inconsistent state-by-state regulations create uncertainty for companies operating nationally. A patchwork of AI laws – especially ones that are not pro-growth – could discourage investment and innovation in a state that does not have thoughtful regulations.

For instance, two slides from a May 30, 2025, 340-page slide deck published by renowned investor and researcher Mary Meeker and her colleagues at BOND show the growing importance of AI and the intense competition after dollars (Figure 18 and 19). Based upon BOND’s figures, global AI sales may reach (or surpass) $1.8 trillion by 2030, a whopping over $1.6 trillion rise from 2022’s $142 billion. Colorado developers compete for this revenue.

 

Figure 17

In addition to the enormous growth in AI-related product sales, adoption of AI tools to a critical mass of users has been incredibly fast. For instance, the following Figure 19 shows the adoption of products. The advancement of AI may make product adoption much faster than we’ve seen in any previous generation with products potentially reaching 100 million users in potentially less than two months.

 

Figure 18

Bottom Line

The astonishingly fast rise in the usage of artificial intelligence and potential for its impact on jobs, economic growth, competition, and consumers presents a challenging environment for balancing competing interests. 

Colorado is currently a national leader in the tech industry, with technology jobs accounting for 10% of all statewide employment and contributing 20% to the state’s total GDP. Colorado also ranks third in the nation for tech sector concentration, highlighting its importance as a hub for innovation and economic growth. Every job created in the tech sector supports an estimated 2.67 additional jobs across other industries, showing the ripple effect and broad economic impact of a strong tech ecosystem. This makes the health and stability of Colorado’s tech industry essential – not only for short-term economic performance but also for long-term statewide prosperity.  

Methodology

Estimating the Impact of Developers

The impact on developers/builders of AI systems was modeled through an impact from the venture capital lens. As mentioned in the main body of the paper, research from economists at the National Association of Business Economics suggests that after passage of the EU’s GDPR AI regulation, weekly venture capital funding declined, potentially by a minimum of 17%. Using Colorado’s $10,426 per $1 million in GDP of venture capital investment, Colorado would forgo some growth in venture capital investment, mostly in technology related firms. This forgone investment in intellectual property was used as an input into REMI.

Estimating the Impact of HR Deployment

To estimate the impact on deployers of AI, the first step was to manually group industries into one of three classifications: Zero to Minimal Impact, Perhaps HR, Broader/Task Level Impact. 

For the industries classified as having Zero to Minimal Impact, no compliance costs were applied.

For the industries with Perhaps HR impact or Broader/Task Level Impact, an HR impact was assigned using the estimated share of wages devoted to human resources technology multiplied by 10%. The result, about $99 million, was applied as an increase in the cost of doing business within the REMI modeling system.