By Lee Allen Miller, Executive Director
I want to close this series the way I opened it. With a practical framework, not a prediction. Over the last ten weeks we have covered where AI belongs in LPTV operations, traffic and scheduling, news writing, sales, playout, captioning, audience analysis, digital and social, and the risks to manage. The question left on the table is how a specific station should turn all of this into a plan for the next year. That is what this article is for.
A good AI roadmap for an LPTV station has three characteristics. It is concrete about which specific applications to pursue in what order. It is realistic about what the station can actually absorb in a given time frame. And it treats AI adoption as an ongoing capability, not a one-time project. Let me walk through how to build one.
Start with your actual priorities
The biggest mistake I see in AI adoption is starting with what is technically interesting rather than what would actually matter for your station. The right question is not what can AI do. It is what are the three or four things at my station that, if they worked better, would most improve the business.
Those might be operational. Your traffic department is perpetually overwhelmed. Your newsroom is stretched too thin to cover what it should. Your sales team needs more pipeline. Your digital presence is anemic. Your captioning is either missing or expensive. Pick the two or three that actually hurt.
Or they might be strategic. You need to differentiate your station in the market. You want to move into a new content area. You are preparing for the ATSC 3.0 transition. You are trying to grow a specific revenue line. Pick the strategic priorities that deserve real investment this year.
Once you have your priority list, match AI applications to those priorities, not the other way around. An AI capability that does not serve one of your real priorities is a distraction, no matter how impressive it is.
Assess your starting position honestly
Before deciding what to pursue, understand where you are. Three areas matter.
Your team’s current AI literacy. Have your staff used AI tools at all? Are they skeptical, enthusiastic, indifferent, anxious? A team with no AI experience and some skepticism will need different change management than a team that is already using tools informally.
Your current technical infrastructure. What systems do you use for traffic, playout, production, and digital? Which of them have AI capabilities built in or available as add-ons? What would be required to integrate external AI tools into your workflows?
Your current data situation. What data do you have, in what systems, in what condition? AI applications are only as good as the data they work with. Stations with clean integrated data will get value from AI quickly. Stations with messy data will need to invest in data hygiene first.
An honest assessment in these three areas will tell you how much you can reasonably take on in a year and what prerequisite work might be needed.
Build the roadmap in phases
A practical AI roadmap for a typical LPTV station has three phases over a twelve to eighteen month period.
Phase one, which should last about three months, is about establishing a foundation. Pick one low-risk high-value application to start with. Usually something in digital, administrative, or content drafting where the outputs go through human review and the failure modes are contained. Deploy it. Train the team. Develop your first policies. Measure what changes. The goal is to build confidence and competence, not to transform the business yet.
Phase two, which should take another three to six months, is about expanding into operational applications. Based on what you learned in phase one, extend AI use into specific operational areas that match your priorities. Traffic optimization. Captioning. Audience analysis. Sales support. Pick two or three applications and implement them deliberately. Refine your policies based on experience. Start measuring productivity and quality impacts.
Phase three, which should take another six months or more, is about integrating AI capabilities into your core workflows. By this point your team has genuine experience, your policies are mature, and you have a realistic sense of what works. You can make larger investments, consider integrated platforms, and pursue applications that require more infrastructure. This is also when AI should start informing your capital planning for future broadcast infrastructure.
The key is that each phase builds on the previous one. You do not start phase two until phase one is working. You do not start phase three until phase two is solid. The temptation to do everything at once is what produces the failed AI implementations we all hear about.
Budget and resourcing
AI adoption requires real resources, but usually not as much as stations fear. A typical first-year AI investment at an LPTV station might include subscription costs for a few AI tools, some training time for staff, possibly a modest consulting engagement to accelerate policy development, and some time from existing staff for implementation and review.
The dollar cost is often in the low thousands to low tens of thousands per year for the tools themselves. The time cost is usually more significant. Stations that underestimate the time component tend to launch initiatives that never quite get finished. Plan for real calendar hours from real people, and protect that time.
Do not plan to hire specialized AI staff at a small station. The economics don’t support it, and the value is usually in helping existing staff use AI tools well rather than in having dedicated AI specialists. Where external expertise is useful, bring it in as consulting rather than as hires.
Governance and policy
Every station adopting AI needs clear policies. Not a thick manual. A working document that covers a few essential questions.
What AI tools are approved for station use and for what purposes.What categories of information can and cannot be shared with AI tools.What human review requirements apply to AI-generated content before it goes public.What disclosure practices apply to AI-assisted content in different contexts.Who is accountable for AI-related decisions and incidents.
Write these policies, make sure staff know them, and update them as you learn more. A station operating without explicit AI policy is accumulating risk that will eventually surface.
Measuring what matters
Track a small number of specific outcomes rather than trying to measure AI comprehensively. For each AI application you deploy, define what success looks like in specific terms. Hours saved. Error rates. Volume of output. Advertiser feedback. Audience engagement. Whatever is relevant.
Review these measurements at a reasonable cadence. What is working better than expected. What is not delivering. What unexpected effects have appeared. Good or bad. Use the reviews to adjust your roadmap, expand what works, and pull back on what does not.
Staying current
AI capabilities are evolving quickly. What is cutting edge now will be standard in two years, and new capabilities will have emerged that we cannot predict. A static roadmap will be outdated almost immediately.
The practical response is to build ongoing learning into the station’s operations. Regular time for leadership to read, attend industry events, and evaluate new capabilities. Occasional pilot projects to test emerging tools against specific use cases. Conversations with peer stations about what they are trying. Engagement with ATBA and other industry organizations that are tracking developments for members.
None of this takes heroic effort. A few hours a month of deliberate attention, applied consistently, will keep you reasonably current without overwhelming the operation.
Closing thought
AI adoption in LPTV is not about chasing the latest trend. It is about using new tools to do the work we have always done, faster and better, so we can focus our limited human resources on the parts of broadcasting that actually require judgment and relationship. That is the same story as every other operational improvement our industry has absorbed over decades. We have done this before. We can do it again.
The stations that handle this transition well over the next several years will be the ones led by people who take the work seriously, maintain clear values about what broadcasting is for, and adopt new capabilities deliberately rather than reflexively. That describes most of the LPTV owners I know, and it is one more reason I am optimistic about our industry even in a time of significant change.
ATBA will continue to help members navigate these questions. Thank you for reading this series. I hope it has been useful, and I welcome conversations about what it has looked like at your station.
