GEB1432 – Applied Artificial Intelligence (AI) in Business is a 3-credit introductory course on using AI tools in ordinary business work: what current systems can and cannot do, how to apply them to real tasks, and how to evaluate their output responsibly.
It is deliberately an applied course rather than a technical one. It does not require programming or mathematics, and it is not a machine learning course — those live under CAI and CAP prefixes. The question here is the practical one facing every business function right now: given that these tools exist, how do you use them well, and how do you avoid the ways they fail?
Content covers what AI is and is not — terminology, capabilities, and the gap between marketing and reality; types of AI in business use — generative, predictive, classification, and recommendation; large language models — how they behave, what they are good at, and their failure modes; prompting — task framing, context, examples, iteration, and evaluation; AI for written communication — drafting, editing, summarizing, and adapting register; AI for analysis — data summarization, exploration, and reporting; AI in marketing — content, segmentation, and personalization; AI in operations and customer service — automation, chat, and routing; AI in finance and accounting — forecasting, anomaly detection, and reconciliation; AI in human resources — screening, and the specific legal risks there; evaluating output — verification, hallucination, and the limits of fluency; data privacy and confidentiality — what must never be entered into a public tool; bias and fairness in business applications; intellectual property and disclosure; governance and policy — acceptable use in an organization; workforce implications; and an applied project.
This course is a capability rather than a career on its own, and it attaches to nearly every business role:
The realistic framing: employers are not generally hiring for "AI skills" as a standalone qualification at this level. They are hiring people who can do a business job well and use these tools competently and responsibly within it. That is what makes this course valuable alongside a functional major rather than instead of one.
The most immediately consequential item, and the one that causes real academic integrity cases. Policies differ by institution, by instructor, and by assignment, and a use that is encouraged in this course may be prohibited in your English or history class the same term.
Do three things. Read the syllabus statement for every course, and where none exists, ask before using AI rather than after. Follow the disclosure requirement exactly — many instructors permit AI use with attribution and treat undisclosed use as the violation. And understand that AI-detection tools are unreliable in both directions: they produce false positives, particularly for multilingual writers, and false negatives routinely. That cuts both ways — do not rely on a detector to clear you, and if you are wrongly accused, ask what evidence beyond a detector score exists. Keeping drafts and version history is the most effective protection available to you.
The single most important professional habit this course should build. Language models produce text that is confident, well-structured, and plausible regardless of whether it is true. They fabricate citations, invent statistics, misattribute quotations, and state outdated information in the present tense — and because the prose is polished, the errors do not look like errors.
The failure is worse in exactly the situations where students rely on it most: unfamiliar topics, where you lack the knowledge to notice the mistake. So the rule is verify anything you will act on or pass along, especially names, numbers, dates, citations, legal and regulatory claims, and anything about a specific company or product.
The professional consequence is real: lawyers have been sanctioned for filing briefs citing cases that did not exist, and businesses have published fabricated figures. You remain responsible for what you send under your name, and "the AI said so" has no standing with a client, a regulator, or an employer.
A specific, avoidable, and career-damaging mistake. Data entered into a consumer AI service may be transmitted to and retained by a third party, and depending on the terms may be used to improve the service. Employees have leaked source code, financial results, and customer data this way, and several large firms restricted or banned public tools for exactly that reason.
What must not go into a public tool: customer and employee personal data, health information, financial and payroll records, unreleased financials, source code, contracts, legal matters, and anything under NDA. In regulated contexts the exposure is legal as well as commercial — HIPAA, FERPA, and financial privacy obligations do not pause because a tool was convenient.
What to do instead: use the enterprise or institutional version if your organization provides one, since those typically carry contractual data protections; anonymize before asking; and read the actual terms on retention and training use rather than assuming. Assume anything you paste could become public, and decide accordingly.
Worth singling out because it is where business use most reliably becomes a legal problem. AI screening and assessment tools have produced documented discriminatory outcomes, and because they learn from historical decisions they can reproduce past patterns at scale while appearing neutral.
The points that matter for a business student: employment discrimination law applies to the outcome, not the intent, so a tool that disadvantages a protected group creates exposure even if no one intended it; "the vendor said it was validated" is not a defense, since the employer remains responsible; disparate impact can arise from proxies that correlate with protected characteristics; and regulation is actively developing, with some jurisdictions requiring bias audits and candidate notification. Verify current requirements rather than relying on a textbook, and treat any tool that influences hiring, promotion, lending, or housing decisions as high-risk by default.
The practical technique, framed usefully. Effective prompting is not a set of magic phrases; it is specifying a task clearly enough that it can be completed — which is the same skill as briefing a new employee or writing a good requirements document.
What reliably improves output: state the role and audience; supply the context and source material rather than expecting recall; give an example of what good looks like; state constraints — length, format, tone, what to exclude; and iterate, treating the first response as a draft to critique rather than an answer.
Two habits worth building. Ask it to show its reasoning or list its assumptions, which frequently exposes where it has gone wrong. And notice when a task is genuinely unsuited to these tools — work requiring current facts you cannot supply, genuine judgment about people, or accountability you cannot delegate. Recognizing that boundary is more valuable than any prompting technique.
Realistic framing for a course whose specific tools will date within a year or two. Model names, interfaces, capabilities, and pricing change continuously, and a course built around one product teaches something with a short shelf life.
What persists: knowing what these systems are fundamentally good and bad at; the verification discipline; the confidentiality rules; the ability to judge whether a proposed application is worth doing; and the ethical and legal framework, which is developing but not arbitrary. A graduate who has those adapts to whatever tool arrives next; one who memorized a particular interface starts over. Treat the specific tools as examples and the judgment as the content.
GEB1432 sits in the general business core alongside GEB1011 (introduction to business), GEB2430 (business ethics), and GEB3213 (business writing), with GEB2351 (international business practice firm) and upper-division GEB4891 (strategic management) elsewhere in the sequence.
Note the prefix distinction, which predicts the course's approach: GEB is general business, so this course is applied and non-technical. The technical treatment of the same subject sits under CAI (artificial intelligence, e.g. CAI4002 Introduction to Artificial Intelligence, which requires data structures and statistics), CAP (applications including machine learning), ISM (information systems management), and CTS (implementation and tools). These are genuinely different courses and do not substitute for one another — choose by what you want to be able to do.
Worth stating precisely, because the numbering is often misread. In Florida's Statewide Course Numbering System the first digit denotes the year in which the course is normally offered — 1 for the first year, 2 for the second, and so on — not how well it transfers. 1000- and 2000-level courses transfer transparently between Florida public institutions, and 3000 to 4000 transfers without difficulty since both are upper division. The boundary that matters is lower division to upper division: taking a 2000-level course toward a 3000-level requirement is the problematic step. PSAV (0-level) courses do not transfer as college credit at all; that pathway runs through articulation agreements instead.
Separately, SCNS equivalency is keyed to the course number. A program requiring a specific number is satisfied by that number from any participating institution; a different number with similar content still transfers as credit, but the receiving program decides whether it fills that requirement or counts as elective. That is a curriculum question for an advisor, not a barrier to the credit transferring.
Generated September 1, 2026 · Updated September 1, 2026