Free BCS Foundation Certificate in Artificial Intelligence Practice Questions

Sample Practice Questions

Try these 10 sample questions from the BCS Foundation Certificate in Artificial Intelligence practice bank. Each includes the answer and explanation. Continue with the interactive practice above to review the full question set.

1. 1An AI research team applies the scientific method when investigating a claim. Which basis for acquiring knowledge characterizes that method in the BCS syllabus?

   1. A.Empirical investigationCorrect answer: A
   2. B.Axiomatic deduction
   3. C.Expert consensus
   4. D.Formal specification

   Explanation: The syllabus calls the scientific method "An empirical method for acquiring knowledge". The other approaches can contribute to inquiry but do not identify the empirical basis described here.

   Reference: BCS Foundation Certificate in Artificial Intelligence v2.0 Syllabus 1.1 — p. 8, objective 1.1(d)

2. 2An AI evaluation team finds systematic errors associated with a nonrepresentative sample. Under the NIST AI Risk Management Framework Playbook, which major category of bias does this illustrate?

   1. A.Systemic bias
   2. B.Human-cognitive bias
   3. C.Computational and statistical biasCorrect answer: C
   4. D.Confirmation bias

   Explanation: NIST states that "Computational and statistical biases can be present in AI datasets and algorithmic processes, and often stem from systematic errors due to non-representative samples." The other options refer to different concepts; confirmation bias is not one of the three major categories named in this NIST passage.

   Reference: AI Risk Management Framework Playbook: Measure — MEASURE 2.11, Fairness and bias

3. 3An analyst groups unlabelled customer records according to similarities without first specifying category labels. Which machine learning task is the analyst performing?

   1. A.Classification
   2. B.ClusteringCorrect answer: B
   3. C.Recommendation
   4. D.Regression

   Explanation: The syllabus describes unsupervised learning using unlabelled data and gives "clustering (grouping data based on similarities)" as an example. The objective is to discover groups rather than assign predefined labels.

   Reference: BCS Foundation Certificate in Artificial Intelligence v2.0 Syllabus 1.1 — p. 16, objective 3.5 guidance

4. 4An analyst receives JSON documents with nested keys and arrays; individual records may contain different fields. Which data category best describes the documents?

   1. A.Structured data
   2. B.Semi-structured dataCorrect answer: B
   3. C.Unstructured data
   4. D.Relational data

   Explanation: The syllabus says semi-structured data "does not follow the tabular structure of a relational database but does have some defining or organisational properties that allow it to be analysed." The fields provide structure without requiring a relational table.

   Reference: BCS Foundation Certificate in Artificial Intelligence v2.0 Syllabus 1.1 — p. 17, objective 4.1(d)

5. 5A project team is preparing a business case for an AI investment. Where in the syllabus's suggested structure should the team compare the costs and benefits of competing approaches?

   1. A.The description of the current state
   2. B.The analysis of options consideredCorrect answer: B
   3. C.The management or executive summary
   4. D.The appendices and supporting information

   Explanation: The BCS syllabus places "Analysis of costs and benefits" under "Options considered". This part of the business case examines competing approaches, while the summary and supporting material have different roles.

   Reference: BCS Foundation Certificate in Artificial Intelligence v2.0 Syllabus 1.1 — p. 21, objective 5.2, indicative content

6. 6An AI chatbot produces convincing conversational behavior. What does the BCS syllabus suggest about the relationship between that behavior and consciousness?

   1. A.The behavior confirms that the chatbot has subjective experience.
   2. B.The behavior may mimic consciousness without establishing subjective experience.Correct answer: B
   3. C.The behavior confirms that the chatbot lacks subjective experience.
   4. D.The behavior measures the intensity of the chatbot's subjective experience.

   Explanation: The syllabus distinguishes "functional capabilities which may mimic consciousness" from "genuine human consciousness". Convincing behavior alone does not establish a subjective experience.

   Reference: BCS Foundation Certificate in Artificial Intelligence v2.0 Syllabus 1.1 — p. 25, objective 6.4 guidance

7. 7An AI history briefing identifies the event regarded as the starting point of AI as a field of practice. Which event should it name?

   1. A.The Asilomar Beneficial AI conference
   2. B.The Dartmouth Summer Research ProjectCorrect answer: B
   3. C.The first AAAI national conference
   4. D.The Dartmouth AI@50 conference

   Explanation: The syllabus states that "The Dartmouth conference which took place in 1956, is considered to be the starting point of AI as a field of practice." The other events came later.

   Reference: BCS Foundation Certificate in Artificial Intelligence v2.0 Syllabus 1.1 — p. 8, objective 1.2 guidance

8. 8An AI team addressing training-data demographic imbalance separates the available population into nonoverlapping demographic groups and independently samples from every group. Which sampling method is the team applying?

   1. A.Cluster sampling
   2. B.Stratified samplingCorrect answer: B
   3. C.Systematic sampling
   4. D.Simple random sampling

   Explanation: NIST defines "Stratified sampling" as dividing the population into mutually exclusive subsets and selecting independent units from each. This gives the team a way to address group representation before model training. Cluster sampling selects clusters; systematic sampling uses an interval; an unrestricted simple random sample does not select independently within each group.

   Reference: AI Risk Management Framework Playbook: Measure — MEASURE 2.11, Fairness and bias

9. 9An AI team starts training with a small labelled dataset and then introduces a larger amount of unlabelled data. Which learning approach does the BCS syllabus describe?

   1. A.Unsupervised learning
   2. B.Reinforcement learning
   3. C.Semi-supervised learningCorrect answer: C
   4. D.Supervised learning

   Explanation: The syllabus explains that semi-supervised learning begins with "a small amount of labelled data" and then introduces "a larger amount of unlabelled data". That combination matches the team's approach.

   Reference: BCS Foundation Certificate in Artificial Intelligence v2.0 Syllabus 1.1 — p. 16, objective 3.5 guidance

10. 10Two active information systems report different birth dates for the same customer. The team has not established which date is correct and wants to assess agreement between the systems. Which data-quality characteristic is in question?

    1. A.ConsistencyCorrect answer: A
    2. B.Uniqueness
    3. C.Accuracy
    4. D.Timeliness

    Explanation: The BCS syllabus describes consistency as "is it free from conflict?" The disagreement between systems is a consistency concern. Without a verified reference, the team cannot determine which date is accurate.

    Reference: BCS Foundation Certificate in Artificial Intelligence v2.0 Syllabus 1.1 — p. 17, objective 4.2(a)