IBM C1000-206: IBM Artificial Intelligence Fundamentals v1 - Associate

Last checked November 10, 2026 against the official IBM exam page.

1. Question 1AI Foundations

   Which two of the following are examples of unstructured data? (Choose two.)

   1. ACustomer emails containing free-form complaints
   2. BRows in a bank's account ledger table
   3. CJPEG photos captured by a security camera
   4. DNumeric columns in a fixed-schema sales database
   Show answer and explanationHide answer

   Correct answer: A, C

   Free-form emails and photos have no predefined schema, so they are unstructured data. Ledger rows and fixed-schema database columns are structured because they follow a predefined table format.

2. Question 2Machine Learning and Deep Learning Fundamentals

   A bank wants to flag each card transaction as fraudulent or legitimate, using a history of transactions that are already labeled with those outcomes. Which approach fits best?

   1. AUnsupervised clustering that groups similar transactions without any labels
   2. BSupervised classification trained on the labeled transactions
   3. CDimensionality reduction that shrinks the number of transaction features
   4. DReinforcement learning that rewards the system for each blocked fraud attempt
   Show answer and explanationHide answer

   Correct answer: B

   Labeled historical outcomes with a yes/no target call for supervised classification. Clustering ignores labels, reinforcement learning learns from rewards in an interactive setting, and dimensionality reduction only compresses features.

3. Question 3Generative AI and Applied AI Concepts

   What does a large language model do at its core when it generates text?

   1. APredicts the next token in a sequence, one token at a time
   2. BStores relational tables and answers SQL queries using indexes
   3. CCompares images pixel by pixel to find duplicate photos
   4. DMonitors network traffic to detect unusual intrusion patterns
   Show answer and explanationHide answer

   Correct answer: A

   A large language model generates text by repeatedly predicting the most likely next token given the tokens so far. Database querying, image comparison and intrusion monitoring are different tasks that a language model is not built to do.

4. Question 4Responsible AI, Ethics, Security, and Governance

   A company plans to use an AI tool that scores résumés and influences which applicants get interviews. Which step should come first before the tool is deployed?

   1. ALaunch it to all recruiters and monitor complaints afterward
   2. BClassify it as high risk and test for bias before rollout
   3. CLet the vendor decide the fairness criteria without company input
   4. DRemove applicant names from the data and skip further review
   Show answer and explanationHide answer

   Correct answer: B

   Recruitment tools that affect who gets hired fall into the high-risk category under the EU AI Act, so the company should assess the risk and test for bias before rollout. Launching first, removing only names, or outsourcing fairness decisions each skip the evaluation that the risk calls for.

5. Question 5AI Use Case Implementation and Business Integration

   A company wants to use AI for the first time. Which candidate project is the strongest starting point?

   1. AA vague goal to transform every business process with AI at once
   2. BA project with no data owner and no agreed success metric
   3. CA well-defined task with historical data and a measurable business outcome
   4. DAn experiment that requires building a new data center before any pilot can begin
   Show answer and explanationHide answer

   Correct answer: C

   A first AI project works best with a clear task, usable historical data and a measurable outcome, since those allow a pilot to be tested and judged. The other options lack a clear goal, an owner or a proportionate scope.