Thank you all for joining me today to talk about how AI and HR intersect. I think this will be an interesting and enlightening conversation! I’m thrilled to welcome Kellis Borek (Vice President of General Counsel and Labor/Legal Services), Caroline Smith (VP of Information Technologies), and Christine Assaf (Senior HR Advisor). Let’s get started!
It’s safe to say the AI we’re seeing today is different than what debuted a few years ago. What is the biggest change you’ve noticed in your day-to-day and how our members are encountering AI?
Caroline Smith (CS): From my perspective, the biggest shift is that AI is showing up before organizations have fully planned for it. The technology is new. It is all over the news and people are experimenting with it before companies might be ready to deploy it. And that is dangerous for data and security and compliance.
That changes the role of leadership. It’s no longer just about introducing AI tools, it’s about catching up with governance—setting expectations, putting basic policies in place, and deciding how AI fits into the culture of the business.
Christine Assaf (CA): In my day-to-day, as an HR Advisor, I’ve noticed members are using AI to research their questions to us at Archbright or for policy creation—we see this in handbook reviews as well. Also, their employees are using AI to research laws and regulations, as well as to draft their responses to HR. This can get burdensome in many ways for us and for the member—AI’s responses can sometimes be lengthy and nuanced and in some cases the responses aren’t entirely accurate.
Kellis Borek (KB): I agree with Caroline and Christine: the "cat is out of the bag” for employers. AI adoption is already happening organically, often ahead of formal strategy, procurement, training, or risk controls. That makes the issue less about whether employees will use AI and more about whether the organization will govern it intentionally.
AI is rapidly evolving, is it worth employers taking the time to put out an AI policy when things are likely to change again before the year is out? How can HR folks stay on top of things without it becoming their full-time job?
CS: Absolutely. Historically, we'd review most IT policies annually. With AI, we've already looked at ours multiple times this year. The technology, the legal landscape, and best practices are changing so quickly that organizations need to be prepared to review and update their AI policies more frequently. I think of it as a living document—it provides the guardrails employees need today while giving the organization flexibility to adapt as AI continues to evolve.
Just as importantly, it's not enough to publish a policy and move on. Organizations need to train employees on what the policy means in practice—what they can use AI for, what information should never be shared, and where human review is still required. As the technology changes, that training needs to evolve right alongside the policy.
The policy sets the guardrails, but training is what helps employees use AI confidently and responsibly.
CA: I’d strongly recommend a policy and yes, that means having to update it as changes come by. As Caroline previously said, the restrictions and usage of it MUST be outlined. Policy changes aren’t new to HR, and when it comes to technology it always moves in a rapid pace. That said, HR needs to ensure they are informed and working closely with IT and other stakeholders around AI policy in the workplace. The best way to stay on top of it is to have other experts—like Archbright, to help!
CS: And, as for staying on top of it without making it a full-time job, I don't think HR should feel like they need to chase every new AI tool or headline. The key is partnering with IT, Legal, and business leaders, establishing a small governance group, and setting a regular review cadence. That way, no one person carries the burden, and the organization can stay informed without becoming overwhelmed.
Related mozzo™ resource: Artificial Intelligence (AI) Readiness and Policy Development Checklist
We know that AI output is not always 100% accurate, what can employers do to mitigate errors and misinformation that may come from AI generated work within their organization?
KB: AI outputs or responses are still presumptively unverified and should not be relied upon without human review. Organizations should start with a clear AI use governance rule/plan that clearly sets expectations about AI content being validated by a human. Additionally, organizations should not use AI for risky categories such as legal advice, employment decisions, financial disclosures, etc. These restrictions can be addressed in an AI policy.
CS: Keeping people in the loop is key. AI saves a lot of time and is great for things like drafting, writing, and improving grammar—but it still needs human review. It’s also important to be aware of bias. AI can reflect patterns from its training data, so we’re not just checking for accuracy—we’re also checking for fairness.
The simple rule I share is: AI can draft, but a person owns the final answer—especially when it impacts decisions, data, or people. Like Kellis said, your AI policy should cover this.
CA: I agree. And check your sources! And build your AI agent to draw from only sources you rely on— government websites, state laws, etc.
Related mozzo™ resource: Artificial Intelligence (AI) Usage Policy
There are some data and privacy concerns related to AI being given total access to workflows. Instances have come up where colleagues might be able to access each other’s private information or documents that otherwise shouldn’t be accessible. On the other side, AI vendors might collect, sell, or mishandle data from their customers, compromising private business records. How can employers take consideration to prevent this sort of data breech inside or out?
CS: AI creates new privacy and security risks, which is why employers need clear policies, employee training, and strong governance. Before deploying an AI tool, organizations should understand what data it can access, how that data is stored, and how the vendor protects it. Organizations should also follow the principle of least privilege—employees and AI tools should only have access to the information necessary to do their jobs.
At the end of the day, AI shouldn't be an exception to your security program. The same privacy, security, and access controls you expect for any business-critical system should apply to AI as well.
KB: As a follow up to Caroline, many employers might focus on whether AI can replace work and not whether AI can be trusted with employer data. Ironically, a single AI related privacy breach could cost more financially and reputationally than the labor savings generated by AI. Along with the principle of least privilege, an employer should review who has permissions prior to deploying AI. For example, an employee might ask AI about compensation trends, but get info from payroll files or privileged documents the employee would never have the right to open directly. Conducting a permission audit and removing legacy permissions is one step to take before deploying AI internally.
How can employers go about educating themselves on AI Tools and what to look for in terms of data security, the learning model, the accuracy of output, and what works with their systems?
CS: You don't need to become an AI expert, but you do need a process for evaluating AI tools. Start with the business problem you're trying to solve, then involve IT, Legal, Security, and the business stakeholders who will use the tool.
When evaluating AI solutions, employers should ask a few key questions: What data does the tool access? How is that data protected? Is customer data used to train the model? How accurate are the outputs? And how well does it integrate with our existing systems?
My advice is to start small and run a pilot with a very specific use case. Don't just give people a new tool and hope they find value. Pick a specific task—drafting job descriptions, summarizing meeting notes, or answering employee questions—and measure the results. That makes it much easier to evaluate whether the tool is delivering real business value before expanding its use.
Related mozzo™ resource: Artificial Intelligence (AI) in the Workplace
AI and Copyright laws have been a hot topic lately as well. Whether it’s training AI on copyrighted materials or companies trying to copyright output from an AI tool we’re seeing a lot of questions and court cases coming up around it. What is the current best practice for using an AI tool on something you intend to copyright in the future?
KB: Employers should seek legal advice on copyright practice but some general guidance recommended by the US Copyright office are: a human should create the concept and make decisions. AI can be used to brainstorm, outline and/or to create rough drafts. Keep records of the creative process such as outlines, drafts, note and prompt history. Copyright protection depends on human authorship. Oh, and don’t put anything you wish to copyright or trademark in the future in AI unless you can completely confirm the platform won’t share it with others.
The news has been buzzing with Pope Leo’s publication of his first encyclical about safeguarding the human person in the time of artificial intelligence. Some people are saying this may be grounds for a religious exemption to AI use at work. One woman has won such a case citing environmental and ethical objections. Should employers prepare themselves for similar cases?
KB: Yes this is an emerging issue. It is not yet a broad new legal right to refuse AI. The woman in the case was granted an accommodation from her employer - it did not create legal precedent. But employers should expect to see more of these requests: e.g., religious accommodation to avoid AI use. This type of request would be governed by existing Title VII and religious accommodation law. If an employee can show a sincerely held religious belief and that conflicts with workplace requirement employers must engage in interactive process for a reasonable accommodation unless it creates an undue hardship. For now, employers should have this on their radar and expect to see these types of requests increase, especially with media attention on the pros and cons of AI. Bottom line, treat this type of request just like any other religious accommodation request.
CA: I think it benefits employers to allow the exemption where it doesn’t impede their essential job functions or create an undue burden, just like employers should always do with any reasonable accommodations.
Related mozzo™ resource: Religious Accommodation Policy
There are few jobs AI isn’t looking to optimize or take over—and in the world of HR there are tools on the market to streamline recruiting, payroll, and more. What do you think AI means for the future of Human Resources as a career?
CA: I think as with every technology tool introduced in HR—it will provide the opportunity to make us faster. That said, it won’t replace us. HR may elevate as it already has away from the practical and into the interpersonal or strategic. Where we used to spend so much time going through physical timecards, digital payroll processing has freed up time for higher level HR. I think AI will provide the same. The speed should free up more time for us to do more of what we’re good at—employee relations, interpersonal needs, culture, etc.
CS: I agree with Christine. I think current professionals (including HR professionals) who will be most successful aren't the ones competing with AI—they're the ones learning how to leverage it effectively. Just as we've adapted to email, cloud computing, and collaboration tools over the years, AI is becoming another workplace technology that professionals need to understand and incorporate into their daily work. Thinking outside the box on how to use these tools will be key.
CA: There’s a really great study that just was published in Washington Post that identified the jobs vulnerable to being replaced by AI. And they determined it based on two things: 1) whether or not the field is exposed to AI and 2) whether the role is adaptable to AI. Jobs that are tangible – like nurses, teachers, firefighters are low in vulnerability. But those in roles with office work or digital fields are exposed and highly vulnerable. But if they adapt to using AI tools, there’s less vulnerability of replacement.
We’re seeing an increase in layoffs that are “AI-Related”. As of June 2026, it was the top cited reason for corporate workforce reductions. What should employers who are considering a workforce reduction in the name of AI consider before turning to layoffs?
KB: A significant risk arises when AI related layoffs or restructuring disproportionately affects protected groups based on age, race, sex disabilities etc. Thus, as in any layoffs, AI driven or not, it’s important to conduct an impact analysis before finalizing selections. As part of this process employers should document a legitimate business case for each layoff. Employers should never rely exclusively on AI to make a layoff decision. As to employees chosen for layoffs, AI cannot replace institutional knowledge, customer relationships, labor relations experience and leadership skills.
CA: I hope they’ve done their due diligence and checked the quality of their work! We’re seeing companies regret layoffs—motor company Ford just rehired its engineers thinking AI could replace and quickly learned that wasn’t the case. Layoffs and rehiring are costly!
So, I’d encourage employees to view adaptability as a way to maintain their role relevancy. And for employers, whether they have enough evidence and data to prove its profitable or sustainable!
CS: Before turning to layoffs, employers should ask whether they're trying to reduce headcount or improve productivity. In many cases, AI can help employees eliminate repetitive tasks and focus on higher-value work without reducing staff.
I also think it's important to invest in training and give employees the opportunity to adapt. AI is changing jobs, including my own, but that doesn't automatically mean fewer jobs. Often, it means people need new skills and new ways of working.
Related mozzo™ resource: Layoffs & Furloughs
What considerations can employers take to balance using AI tools and maintaining the culture and morale of their teams?
CA: Let’s all take a step back and remember the 1990 dot com era. Many things emerged, but very few remained. Remember Zappos? Remember AOL? Things moved fast then and that level of speed and growth wasn’t sustainable. So we have to remind ourselves that things are fast right now, but dust will settle. There's always change, but some things remain constant. AI as a tool may be here to stay but having a capable workforce will always be needed and a more important investment.
Leaders and decision makers should remember that employees may be scared and anxious, and hesitant to adapt, so proceed forward with caution, optimism, and empathy. Reminders that employees are your most important resource and providing them with training and support to integrate these tools is key.
CS: Organizations need a strong culture built on trust. Employees need to understand why AI is being introduced, how it will be used, and what it means for their role. If people see AI as something being done to them, they'll resist it. If they see it as a tool that helps them be more successful, adoption is much more likely.
It also requires embracing new ways of thinking. Organizations need to give employees the time and space to experiment, learn, make mistakes, and try again. Not every AI pilot will be successful, and that's okay. Some of the biggest lessons come from the projects that don't work as expected.
The goal should be to create an environment where employees feel safe exploring new technologies, sharing what they've learned, and continuously improving. That's how organizations build both innovation and trust.
KB: Ditto regarding Caroline and Christine's responses. I would add transparency as being one of the most important factors in whether an employer’s AI strategy succeeds or fails from its people’s perspective. Clearly and specifically communicate the “why” behind introducing AI to the organization – employees need to hear more than “we are introducing AI to improve efficiency” and do not make overoptimistic assurances that later prove inaccurate, such as “AI will never affect jobs”.
Thank you all for joining us to talk about the Impact of AI on HR. And thank you to everyone who is reading along with us! If you have questions about the impact of AI on HR that we didn't cover here, email them to info@archbright.com and we may answer them in our next Q&A!
Need additional support with your AI policy? Want to talk to an HR Advisor about how to address AI Questions from your employees? Archbright members have unlimited access to resources in mozzo including our Advisor Chat staffed by real Human Resource experts. Members at the silver tier and above can also contact our legal team for advice.
Learn more about Archbright Membership →