As Colorado refines its approach to regulating artificial intelligence in the workplace, employers and their advising attorneys are confronting a fast-moving legal landscape.
New research from Littler shows that 84% of U.S. employers expect AI-related policy changes to impact their businesses in the next year. This is double the share noted in the same in Littler’s 2025 survey (42%), underscoring how quickly the issue has moved from emerging concern to top priority.
A Broader, Blurrier Framework
Changes this year in Colorado law are reshaping compliance obligations.
“The new law applies to ‘automated decision-making technologies,’ which is a much broader category than ‘artificial intelligence,’” said Zoe Argento, co-chair of Littler’s Privacy and Data Security Practice Group. “It pulls in technologies that have been around for quite some time and that some companies may not think of as AI.”
She added, “On balance, compliance is easier under the new law, but it’s still challenging.”
Employers must now determine whether a tool falls into a clerical carveout or qualifies as regulated decision-making technology. “The line will be blurry for some time,” Argento said.
Compliance Pressure Around Transparency
Even before the emergence of AI-specific regulation, employers faced legal exposure under longstanding anti-discrimination laws. 2026 legislation layers additional requirements, particularly around transparency.
Colorado’s new Automated Decision-Making Technology law includes notice obligations when automated tools contribute to adverse employment decisions. “That dramatically increases the pressure on employers to have internal policies,” Argento said.
Littler’s survey found 55% of employers currently have formal review processes for AI tools or use cases, while 68% have a formal policy governing AI use in the workplace.
The Human Element
One of the most consequential questions is how much human involvement is required when employers rely on automated tools.
Many laws, including Colorado’s, hinge on whether technology “materially influences” a consequential decision, a standard that remains open to interpretation. “That is one of the toughest questions in these AI laws,” Argento said.
In practice, ambiguity is pushing employers toward more hands-on oversight. “It does seem to imply human review of each decision that qualifies,” she said.
Maintaining human involvement can help companies avoid triggering certain regulatory requirements, too. “If they have adequate human intervention, most of the new AI laws don’t apply,” Argento noted.
Data Privacy Risks Expand Alongside AI Adoption
As companies adopt AI tools to increase efficiency, they confront less visible risks tied to data collection.
Employers may not fully understand what information their systems are gathering, especially when vendors provide limited transparency.
“It can be intimidating for companies when they’re presented with these tools,” Argento said. “But there are questions they can ask, starting with simply, ‘What data is being collected?’”
Investigating how data is used is critical, as some tools process sensitive information like biometric data. For example, systems used to evaluate video interviews may analyze facial features or voice patterns, potentially triggering legal obligations.
For employers and those advising them, the takeaway is straightforward: “You have to understand the technology carefully before you implement it,” she said.
A Patchwork of Laws
Complicating compliance is the growing patchwork of state laws governing AI and data privacy.
While several states and localities are developing their own frameworks, multistate employers are left navigating inconsistent requirements.
“Industry and big multistate companies would prefer a national framework,” Argento said. “The chances of that passing anytime soon are very, very low.”
In the absence of federal legislation, California is expected to play an outsized role in shaping how companies respond. “Where California goes, the rest of the country tends to go,” she said, noting that most large employers will build compliance programs around California’s standards.
Strategy Over Reaction
“AI is going to come into the workplace,” Argento said. “Companies should be strategic about adoption rather than being reactive.” Being strategic involves evaluating tools carefully, implementing governance structures and training employees to use AI effectively.
She continued, “I think employers often make a mistake when they assume that AI is going to replace people. AI plus people is much more powerful than AI alone.”