<body><div><figure><img src="https://img.particlenews.com/image.php?type=thumbnail_580x000&amp;url=10grDc_1H3mv3zy00" alt="https://img.particlenews.com/image.php?url=10grDc_1H3mv3zy00" width="560" height="300" ><figcaption>Athletic departments nationwide are testing AI tools to separate them from expensive novelties and determine what products can help their programs.</figcaption></figure>At Plymouth High School in Canton, Michigan, Chaka Johnson’s artificial intelligence pilot does not have a steering committee, a rollout calendar or a major software contract. In fact, it only has one daily user: Johnson.The athletic director tests AI tools on his own workload before asking anyone else in the department to use them. “Tasks that used to eat an hour of my evening, like drafting a clear parent email or building a game-day graphic package, now take minutes,” Johnson said. “That time goes back to being present with our coaches and kids.”Utilizing AI, he drafts communications, builds meeting agendas, creates social media content and produces game-night graphics. He has also used AI to review policies, explore facility layouts and send rapid notifications during an emergency facility change.That hands-on approach points to the central question athletic departments should answer before buying an AI product: Does it solve a recurring problem under the conditions in which the department actually works?<h2>Start with the work</h2>Product demonstrations can make almost any platform look indispensable. Johnson begins instead with a slow, repetitive or inconsistent task and tests whether AI improves it during a real school day.<blockquote>“If I cannot name a specific task in my department that is slow, repetitive or inconsistent, I do not need the tool, no matter how good the demo looks.” — Chaka Johnson, athletic director, Plymouth High School</blockquote>“I start with the problem, not the tool,” Johnson said. “If I cannot name a specific task in my department that is slow, repetitive or inconsistent, I do not need the tool, no matter how good the demo looks.”That test should use the department’s actual workflow, not a vendor’s ideal scenario. Scheduling conflicts, eligibility questions and emotionally charged parent communications expose weaknesses that a polished demonstration might hide. A platform offers little value if its intended users find it too cumbersome for their work.“Pilot with your own data and your own people before you spend anything,” Johnson said. “Ask the vendor how the tool handles the messy, specific situations you deal with daily.”<h2>Define success before the test</h2>A pilot needs a baseline and a finish line, and Johnson measures AI’s value through time, clarity and consistency.A parent message that once required 45 minutes of drafting and revision might take 10 minutes. A clearer message can also reduce confused replies and repeated conversations. Meanwhile, recurring documents, including coach evaluations, can follow a more consistent structure without surrendering the athletic director’s voice.Those observations can become a simple scorecard. Before testing begins, record how long the task takes, how often errors or follow-up questions occur and whether it is completed on time. Run the pilot through busy weeks and exceptions, then compare the results with the baseline, including time spent on training, editing and oversight.Johnson treats every AI-generated product as a first draft.“I never let it send on my behalf,” Johnson said. “If I read the output and it does not sound like Plymouth Athletics, it does not go out.”<h2>Build knowledge before buying</h2>The University of Toledo provides a larger-scale example. In 2025, its athletics department required AI training for coaches, staff and administrators and made Microsoft Copilot available to university employees. The goal was to establish a common level of understanding before expanding AI use.Under then-athletic director Bryan Blair (now director of athletics at Syracuse), staff used Copilot and ChatGPT for tasks such as drafting scouting reports and donor thank-you notes and creating concepts for facility renovations. Blair calculated that if each employee saved one hour per week, the department would gain capacity comparable to approximately 15 additional staff positions.The lesson for smaller departments is not that everyone needs the same platform. Staff members need enough knowledge to evaluate claims, recognize weak output and determine where automation belongs. Athletic directors who have used general-purpose AI tools are better prepared to question a vendor about accuracy, workflow, training demands and total cost.<h2>Treat a free pilot like a purchase</h2>Efficiency cannot be the only standard when AI systems touch student information. Athletic departments may hold health information, medication records, academic performance, eligibility details, behavioral information and, in some programs, biometric or location data.Linnette Attai, an independent student data privacy consultant and president of PlayWell LLC, a privacy and compliance consultancy specializing in youth and education, noted athletic directors should not make privacy, security and contracting decisions alone. The institution’s chief technology officer, chief information officer, chief privacy officer or another designated specialist should conduct the formal review.“Do not go it alone,” Attai said. “Build a partnership with the person at your institution who is already responsible for assessing the privacy and security of technology products.”That review is necessary even when a vendor offers a free, short-term pilot. The school should know what personal information the product collects, how it is used and protected, who can access it, how long it is retained and how it will be deleted. The contract should limit the vendor to the approved educational purpose and preserve the institution’s control over inputs and outputs.AI model training deserves specific attention. A company should not use student personal information to improve its general commercial model. Schools also need procedures for accessing, correcting and deleting student information when parents or eligible students exercise their rights. Depending on the state and the proposed use, biometric data may trigger additional consent requirements.“Whether a product is free or paid, the privacy and security requirements remain the same,” Attai said. “Do not be lulled into a false sense of security by thinking a short-term pilot means the risk is low.”<h2>Keep consequential decisions human</h2>Before approval, departments should define what the system may assist with and what it may never decide. AI can organize information, identify patterns and prepare a draft. Decisions that affect safety, eligibility, discipline, employment or a student’s educational opportunities require an authorized person to review the evidence, understand the context and take responsibility for the outcome.The same principle applies to communication. Johnson removes names, grades, medical information, disciplinary records and family contact information before using AI to help draft a sensitive message. He adds the specifics himself and reviews every word before it leaves his office.Purchasing should come only after the department has proved the need, measured the benefit, completed the institutional review and assigned human accountability. Otherwise, an impressive pilot can become an expensive subscription that adds oversight without improving the work.“AI should make you a better decision-maker, not replace the decisions only you can make,” Johnson said. “Those moments require judgment, context and trust, and no tool can hand you those.” ◼︎Keith Loria is a contributing writer at Coach &amp; A.D. and can be reached at [email protected]. </div></body>

Block Source: coachad.com

Misinformation

Clickbait

Old or repetitive news

Explicit content

Location not relevant

At Plymouth High School in Canton, Michigan, Chaka Johnson’s artificial intelligence pilot does not have a steering committee, a rollout calendar or a major software contract. In fact, it only has one daily user: Johnson. The athletic director tests AI tools on his own workload before asking anyone else...

Beyond the pilot program: Separating useful AI tools from expensive novelties

NewsBreak

At Plymouth High School in Canton, Michigan, Chaka Johnson’s artificial intelligence pilot does not have a steering committee, a rollout calendar or a majo

Beyond the pilot program: Separating useful AI tools from expensive novelties

Comments