Michigan has always been a place where industrial change becomes visible before the rest of the country fully understands what is happening. What begins in Michigan’s factories, engineering centers, and supplier networks often becomes a preview of the next American economy. Today, that pattern is repeating itself through artificial intelligence. Manufacturers across the state are already using AI to improve production scheduling, anticipate equipment failures, inspect parts with machine vision, model logistics, reduce scrap, and support engineering work with faster analysis. At the same time, lawmakers, regulators, and business leaders are trying to define the legal guardrails for this new era. The result is a moment of unusual consequence for Michigan. The state is not simply adopting another layer of software. It is deciding how automation, data, labor, safety, and industrial competitiveness will fit together in a manufacturing economy that remains central to its identity and prosperity. ² ³
Any serious discussion of AI in Michigan manufacturing has to begin with the structure of the state’s economy. Manufacturing is not a niche sector in Michigan. It remains one of the state’s most important economic engines, employing hundreds of thousands of workers and supporting an ecosystem that reaches far beyond assembly plants. The state’s own workforce analysis has shown that Michigan’s manufacturing cluster employs nearly 586,900 people and accounts for 14.2 percent of total statewide jobs, with wages that exceed the statewide average. ³ Those figures matter because they explain why AI policy in Michigan will never be only about technology. It will be about wages, job quality, regional investment, workforce mobility, supplier resilience, and the competitiveness of one of the most concentrated manufacturing economies in the United States. In Michigan, debates over AI are debates over the future of middle-class work.
That is also why the legal dimension of AI is so important. Public discussion often treats AI law as though it were a separate field reserved for software companies, start-ups, and national regulators. In practice, manufacturers do not experience the law that way. They experience it through procurement contracts, workplace rules, discrimination claims, cybersecurity policies, product liability exposure, recordkeeping duties, and board-level decisions about risk. A manufacturer deciding whether to deploy AI for shift allocation, defect detection, design optimization, predictive maintenance, or recruiting is not simply adopting a tool. It is making a series of legal judgments, whether or not it recognizes them as such. The more deeply AI moves into business operations, the more those judgments become strategic rather than merely technical. ⁴ ⁵
Michigan is still in the early stages of building a formal AI-specific legal framework, but that does not mean the state is legally unregulated. In fact, one of the most important realities for manufacturers is that AI is arriving faster than comprehensive statutes. Michigan has seen AI-related legislative activity, including House Bill 4668 in the 2025-2026 session, which would create an “Artificial Intelligence Safety and Security Transparency Act” aimed at large developers of foundation models and would require safety and security protocols for managing critical risks. ¹ That bill is notable not simply because of its content, but because it signals that AI regulation is no longer theoretical in Michigan. The state legislature is actively considering how to govern AI systems, risk management, disclosures, and accountability. Even where proposals do not immediately become law, they establish a policy direction and reveal what state officials increasingly view as legitimate areas of oversight.
For manufacturers, the significance of proposed legislation like House Bill 4668 lies in the broader signal it sends. Michigan is beginning to move away from the assumption that AI should be governed only through general existing law and toward the idea that some AI systems may warrant rules aimed specifically at their scale, risk, and potential harms. ¹ That matters for the manufacturing sector because industrial AI is likely to sit at the intersection of physical risk and digital decision-making. A foundation model used in office work raises one set of concerns; an AI-enabled system used to support industrial maintenance, industrial design, robotics, safety analytics, or quality assurance can affect physical processes, worker behavior, and production outcomes. As lawmakers become more attentive to high-impact AI, manufacturers should expect closer scrutiny of how these systems are selected, tested, documented, and monitored.
Even before Michigan adopts a broad AI statute, existing legal obligations already shape what companies can and cannot do. That point is especially important because many businesses assume that if no AI-specific law expressly prohibits a practice, the practice is legally safe. That is not a sound assumption. Federal employment law still applies when employers use AI in recruiting, hiring, evaluation, promotion, pay, or workplace monitoring. The U.S. Equal Employment Opportunity Commission has made clear that employers remain responsible when AI systems discriminate on the basis of protected characteristics and that existing law may also require reasonable accommodations when automated systems disadvantage workers because of disability, religion, pregnancy-related limitations, or similar protected conditions.⁵ In other words, the use of an algorithm does not dilute the employer’s legal duties. It can actually create new paths to liability if managers rely on automated systems without validating their fairness and suitability.
This point deserves special emphasis in Michigan manufacturing because manufacturers are increasingly interested in AI not only for production, but also for labor management. In a tight labor market, companies understandably want better tools for recruiting skilled workers, forecasting attrition, matching employees to shifts, identifying training needs, and evaluating performance. Yet the legal risk grows when those tools move from advisory use into consequential decision-making. If a plant adopts an AI system to screen resumes for maintenance technicians, score recorded interviews for production supervisors, predict which employees are least likely to stay, or recommend promotions on the basis of digital performance data, the company must ask whether the system has a disparate impact, whether it can be meaningfully audited, and whether managers understand when to override the machine.⁵ The law will not be impressed by a defense that the system was “industry standard” if it produced discriminatory results.
The same is true of workplace surveillance. Many industrial employers are drawn to AI-enabled tools that monitor throughput, location, safety compliance, machine interaction, or even worker movement patterns. Some of these applications may genuinely improve safety and reduce downtime. But the legal and ethical questions become sharper as surveillance becomes more granular and more predictive. If workers believe they are being constantly scored by opaque systems, the result may not be efficiency. It may be distrust, labor conflict, reduced morale, and legal challenge. The EEOC’s guidance is therefore relevant not only to hiring but also to ongoing employment decisions, including systems that monitor keystrokes, movement, facial expressions, or productivity signals. ⁵ For Michigan manufacturers, the practical lesson is simple: every workplace AI tool should be treated as both an operational instrument and an employment-law exposure.
The legal future of AI in manufacturing is also being shaped by risk-management standards rather than only by statutes. This is where the National Institute of Standards and Technology has become highly influential. NIST’s AI Risk Management Framework is voluntary, but it has quickly become one of the most important reference points for organizations that want to adopt AI responsibly. NIST describes the framework as a way to improve the incorporation of trustworthiness considerations into the design, development, use, and evaluation of AI products, services, and systems, and it has expanded its guidance with a generative AI profile and a 2026 concept note addressing trustworthy AI in critical infrastructure.⁴ For manufacturers in Michigan, that framework is more than a policy document. It is increasingly a practical blueprint for internal governance.
What makes the NIST framework especially relevant to manufacturing is its emphasis on governance, measurement, and continuous management rather than one-time compliance. Industrial settings are dynamic. A model that performs adequately in testing may drift when production inputs change, when suppliers change, when workers alter routines, or when the system is repurposed beyond its original scope. NIST’s approach encourages organizations to think in terms of lifecycle risk: what the model was designed to do, how it is used in practice, how performance is monitored, how harms are identified, and how accountability is assigned. ⁴ This map well onto manufacturing operations, where quality systems, safety systems, and continuous improvement cultures already exist. The most forward-looking Michigan manufacturers will not treat AI governance as an external legal burden. They will incorporate it into the same disciplined operating logic that already governs plant safety, lean production, and supplier management.
That shift from compliance to governance could become one of the defining competitive advantages in Michigan. The state’s manufacturers are unusually well positioned to operationalize AI responsibly because they already understand process discipline. They know how to validate equipment, document deviations, conduct root-cause analysis, trace failures, and standardize corrective action. Those same habits can be adapted to AI deployment. A manufacturer that can show why a model was adopted, what data it relied on, what risks were considered, how human oversight works, what metrics are monitored, and how concerns are escalated will be in a much stronger position than a company that buys AI through a vendor and assumes the problem is solved.⁴ ⁵ In future litigation, regulatory inquiries, customer audits, and internal investigations, documentation will matter. Michigan’s industrial culture gives many firms a head start if they choose to use it.
The economic stakes are high enough that Michigan’s public policy is increasingly framed in terms of strategic opportunity as well as risk. The state’s AI and Workforce Plan argues that Michigan should lead, not follow, in integrating AI into workforce development and economic growth, and it explicitly presents AI as a transformative force affecting how people work, learn, and deliver services. ² The plan connects AI adoption to broader statewide goals around skills, opportunity, and business growth, reflecting a belief that the right response is not to resist AI but to shape its use so that economic gains are broadly shared. This matters for manufacturing because the sector will likely be one of the primary arenas in which that policy ambition is tested. If Michigan can modernize factories while preserving job quality and expanding training pathways, it may demonstrate a model that other industrial states will try to copy.
The plan is also notable because it rejects a purely defensive posture. It does not present AI as something that happens to workers from the outside. Instead, it frames AI as a force that can be directed through strategy, infrastructure, education, and public coordination. ² That framing is especially important in a manufacturing state where the politics of automation have always been emotionally charged. Workers have long heard promises that new technology will create better opportunities later, while often seeing immediate disruption first. Michigan’s challenge is therefore not merely technical adoption. It is institutional credibility. Workers, unions, educators, and employers need to believe that AI deployment will be accompanied by real investment in skills and transition support. Without that trust, even sensible innovation strategies can encounter resistance.
Michigan’s manufacturing workforce profile makes that challenge more urgent. The state’s manufacturing analysis shows that the sector has a relatively high share of workers between ages 45 and 64, while many key occupations require a high school diploma or equivalent plus on-the-job training rather than advanced formal degrees. ³ That combination has major implications for AI policy. It suggests that the future of manufacturing in Michigan will not be determined solely by elite data scientists or software engineers. It will depend on whether experienced technicians, machine operators, maintenance staff, logistics workers, and frontline supervisors can adapt to AI-enhanced systems without being pushed aside. AI law and policy in Michigan will therefore be judged in part by whether they support transition, training, and human capability rather than simply rewarding capital investment in automation.
This is why workforce development should be understood as part of AI governance, not as a separate social policy. A manufacturer that deploys AI without investing in worker comprehension creates both operational and legal risk. Workers who do not understand why an algorithm is making recommendations are less likely to trust it, less likely to use it correctly, and less able to detect when it is wrong. Supervisors who rely on AI outputs without understanding the system’s limits may unintentionally create safety issues, quality failures, or discriminatory outcomes. In contrast, a company that treats workers as participants in implementation rather than passive subjects is more likely to build durable adoption. Michigan’s AI and Workforce Plan is significant because it recognizes that the future of AI is not just about invention. It is about institutional capacity across the labor market. ²
The legal environment will probably continue to develop unevenly, with some rules arising through legislation and others through enforcement, litigation, procurement standards, insurance requirements, and customer expectations. That is often how transformative technology becomes regulated in practice. Manufacturing companies should not assume that regulation will arrive only through one comprehensive AI law. Instead, they should expect a layered environment in which different issues are governed through different channels. Employment decisions may be shaped by anti-discrimination law. Safety-sensitive systems may be influenced by standards and product-liability reasoning. Contracts may impose audit rights or security requirements. Customers may ask for evidence of responsible AI use before awarding work. Proposed state laws like Michigan’s House Bill 4668 may add direct obligations over time, while federal guidance and frameworks continue to define best practice. ¹ ⁴ ⁵
That layered environment will be especially important for suppliers. Michigan manufacturing is famous for large anchor firms, but the state’s industrial strength also depends on thousands of small and midsize companies. These firms often face the hardest challenge because they are under pressure to modernize but have fewer internal legal and technical resources. A large automaker or advanced manufacturer may be able to create an internal AI governance council, retain outside counsel, and conduct vendor due diligence. A smaller tool-and-die shop or precision manufacturer may instead adopt software marketed as simple efficiency improvement. Yet the legal and operational consequences can still be substantial. If a vendor’s system is biased, insecure, unreliable, or poorly documented, the manufacturer using it may bear the practical consequences. For that reason, the future of AI in Michigan manufacturing will depend not only on frontier innovation but also on whether small and midsize firms receive support in evaluating and governing AI adoption. ² ³
This is where Michigan’s public strategy could matter most. The state has increasingly signaled that it wants to create an ecosystem in which businesses can modernize while navigating change more effectively. The AI and Workforce Plan and related state initiatives point toward a more coordinated vision in which workforce institutions, employers, and public agencies work together rather than in isolation. ² If that vision is implemented well, Michigan could distinguish itself by reducing the gap between policy rhetoric and factory-floor reality. A good AI ecosystem for manufacturing would not simply reward deployment. It would help companies understand legal exposure, train workers, adopt sound standards, and avoid the common trap of buying impressive tools with weak implementation discipline.
Manufacturing executives in Michigan should therefore stop asking only whether AI will save labor hours or reduce downtime. Those are fair questions, but they are no longer sufficient. The better question is whether the organization can govern AI in a way that strengthens resilience, trust, and competitiveness over time. That means asking whether the data feeding the system is reliable, whether vendors are transparent enough, whether model outputs can be challenged, whether workers know how decisions are made, whether discrimination risks have been assessed, whether cybersecurity and confidentiality have been considered, and whether the company can explain its use of AI to a regulator, judge, customer, insurer, or union representative. ¹ ⁴ ⁵ In the coming decade, the firms that can answer those questions confidently are likely to outperform those that treat AI as a black box.
This emerging legal landscape may also change the role of lawyers and compliance professionals inside manufacturing companies. Traditionally, legal review often entered late in the process, after operations or procurement teams had already chosen a vendor or defined a project. With AI, that sequence is becoming riskier. Legal counsel must now engage earlier, not to block innovation, but to shape it. The best in-house and outside lawyers will not merely review contract language. They will help define governance structures, documentation practices, escalation pathways, and accountability for AI-enabled decisions. They will work with HR, IT, operations, and executive leadership to translate broad principles into plant-level procedures. In that sense, AI law in Michigan manufacturing will not be a narrow specialty. It will become part of ordinary industrial governance.
There is also a broader political question under the surface of this conversation: what kind of manufacturing future does Michigan want? One path would treat AI primarily as a labor-reduction tool and accept the social consequences as unavoidable. Another would treat AI as a means of augmenting skilled work, improving safety, increasing throughput, and making manufacturing jobs more technically sophisticated and better paid. The legal system cannot fully determine which path firms choose, but it can influence incentives. Laws and standards that reward transparency, accountability, human oversight, and worker access to opportunity can help steer deployment toward augmentation rather than dispossession. ¹ ² ⁴ ⁵ That distinction may sound abstract now, but it will become concrete in the design of recruiting systems, monitoring tools, quality platforms, and production planning software.
Michigan’s history suggests that the state understands the cost of getting industrial transitions wrong. When technological change is managed without regard to workers or communities, the resulting damage can last for generations. But Michigan’s history also shows that industrial reinvention is possible when the state aligns skills, capital, public policy, and business strategy. That is why the current moment is so consequential. AI is arriving just as the state is also thinking about advanced manufacturing, semiconductors, electrification, and supply-chain resilience. ² ³ In that environment, AI law should not be viewed as a brake on growth. Properly understood, it is part of the infrastructure of trust required for long-term investment.
For business leaders, the most prudent posture is neither panic nor complacency. Michigan does not yet have a mature, comprehensive AI code governing every industrial use case, but the legal expectations are becoming clearer. Proposed state legislation shows growing interest in direct AI oversight. ¹ Federal civil rights law already reaches workplace AI. ⁵ NIST provides a credible governance framework that many organizations will increasingly treat as the benchmark for reasonable practice. ⁴ State workforce planning shows that Michigan intends to compete aggressively in an AI-shaped economy rather than retreat from it, ² and the underlying structure of Michigan’s manufacturing labor market makes the stakes impossible to ignore. ³ The convergence of these developments means that the future is not waiting for companies to get ready. It is already arriving through procurement decisions, software deployments, and plant-level experimentation.
The manufacturers that succeed in this environment will likely be the ones that combine legal seriousness with operational imagination. They will move beyond the false choice between innovation and regulation. Instead, they will recognize that the next generation of industrial competitiveness will depend on governing technology well. They will build AI adoption around traceability, documentation, testing, fairness review, worker training, and executive accountability. They will ask more from vendors. They will involve HR and legal teams earlier. They will create channels for employees to question automated decisions. They will understand that public trust, worker trust, and customer trust are all strategic assets. In a state like Michigan, where manufacturing still defines so much of the economy and civic identity, that approach is not just ethically preferable. It is commercially intelligent.
The future of manufacturing in Michigan will not be written by technology alone. It will be written by the interaction of law, labor, management, standards, and public ambition. AI can help Michigan build smarter factories, stronger supply chains, and more adaptive production systems. It can help preserve competitiveness in an era of intense global pressure. It can also create new forms of opacity, inequity, and risk if companies deploy it carelessly or lawmakers respond too slowly. The question, then, is not whether AI will shape Michigan manufacturing. It already is. The real question is whether Michigan will shape AI in a way that reflects the best of its industrial tradition: technical excellence, practical problem-solving, and a belief that economic progress should remain tied to broad human opportunity. If the state can do that, then the next chapter of Michigan manufacturing may be defined not by fear of automation, but by responsible innovation strong enough to support both productivity and dignity. ¹ ² ³ ⁴ ⁵
Contact Tishkoff
Tishkoff PLC specializes in business law and litigation. For inquiries, contact us at www.tish.law/contact/. & check out Tishkoff PLC’s Website (www.Tish.Law/), eBooks (www.Tish.Law/e-books), Blogs (www.Tish.Law/blog) and References (www.Tish.Law/resources).
Footnoted Sources
- Michigan Legislature, House Bill 4668 of 2025, “Artificial Intelligence Safety and Security Transparency Act,” bill text and legislative summary, introduced June 24, 2025. https://legiscan.com/MI/text/HB4668/id/3258161
- Michigan Department of Labor and Economic Opportunity, Michigan’s AI and the Workforce Plan: An Addendum to the Michigan Statewide Workforce Plan, 2025. https://www.michigan.gov/leo/news/2025/05/29/ai-and-the-workforce-plan-will-create-jobs-invest-in-workforce-and-enhance-economic-growth
- Michigan Center for Data and Analytics and Michigan Department of Labor and Economic Opportunity, Michigan Manufacturing Industry Cluster Workforce Analysis, 2023. https://www.michigan.gov/mcda/reports/michigan-industry-workforce-analysis-reports
- National Institute of Standards and Technology, AI Risk Management Framework and related implementation materials, updated through April 7, 2026. https://www.nist.gov/itl/ai-risk-management-framework
- U.S. Equal Employment Opportunity Commission, Employment Discrimination and AI for Workers, April 29, 2024. https://www.eeoc.gov/sites/default/files/2024-04/20240429_Employment%20Discrimination%20and%20AI%20for%20Workers.pdf
This publication is for general informational purposes and does not constitute legal advice. Reading it does not create an attorney-client relationship. You should consult counsel for advice on your specific circumstances.