Are You Ready for the AIGP Exam?

Let's try a practice test

Get your results instantly. Take our 10 practice questions and get an indication of where you stand!

1/10

Practice Questions

Kristin has been tasked with developing a proprietary AI/ML model for her company that will be incorporated into its end-user products. After initial training, Kristin realizes the model produces poor predictive results and fails several measures of accuracy. After reviewing the design of the system, Kristin is convinced that the model has failed to fully capture the complexity of the underlying training data. Which of the following best describes the problem the Kristin has identified?

The model is underfit.

The model is overfit.

The model contains too much entropy.

The model was subjected to data poisoning.

All of the following are sub-dimensions of the Task & Output dimension of the OECD's Framework for the Classification of AI Systems, except:

Which of the following characteristics of AI creates the biggest impediment to redressability by end users?

AcmeAI has compiled a dataset and proprietary algorithm to train an AI/ML foundation model. It then licenses this model to third parties that can incorporate that model into their products and services. BigData Co. relies upon the foundation model from AcmeAI to look for anomalies in large pools of data that it processes. In order to effectively do this, BigData fine-tunes the foundation model to fit its unique use case. BigData produces reports for industry participants about the data that it processes. What role is played by the industry participants that read the reports produced by BigData?

Industry participants are deployers of the AI system.

Industry participants are the users of the AI system.

Industry participants are neither a deployer, developer, or user of the AI system.

If Company X contracts with Company Y to run analytics on data, and Company Y subcontracts with Company Z to store that data, which accurately describes each person's role?

Company X is a data controller, Company Y is a data controller, and Company Z is a data processor.

Company X is a data subject, Company Y is a data controller, and Company Z is a data processor.

All three companies are data processors.

Company X is a data controller, Company Y is a data processor, and Company Z is a data processor.

All of the following factors should be considered in determining whether a human has made a "significant contribution" to an AI-assisted invention, except:

The significance of the contribution.

The quality of the contribution.

The purpose of the contribution.

The addition of something new.

Which of the following is not an explicit transparency requirement mandated by the E.U. AI Act?

Employees must be notified that they are subjected to the use of a high-risk AI system.

Users of high-risk AI systems must be notified if their use of that system results in automated decision making about them.

Users of generative AI systems must be notified that they are using a generative AI system, unless it is obvious to a reasonably well-informed person.

Users of general purpose AI systems must be informed that they are interacting with a foundation model, rather than a more specific, fine-tuned model.

What core function under the NIST AI Risk Management Framework is a cross-cutting function that is infused throughout the other three functions?

Acme Co. was an early adopter of AI. It relies upon third-party vendors to provide foundation models that Acme then fine-tunes to suit its individual needs and business use cases. It has noticed overtime that some of the AI models that it uses will suddenly degrade over a short period of time. Acme discovers this degradation because it regularly conducts tests that are designed to determine whether the AI model performs consistently and accurately when exposed to new data. Which of the following best describes Acme's regular testing?

It is an AI impact assessment.

It is a type of reliability assessment.

It is a type of assessment that only those subject to the E.U. AI Act should perform.

It likely utilizes a confusion matrix.

Which of the following is most accurate regarding the training, validation, and testing of an AI/ML model?

Training, validation, and testing is an iterative process, rather than a strict life cycle process.

Training, validating, and testing a model is done pursuant to standardized measures called benchmarks.

Testing and validation are only designed to measure the accuracy of model training.

Testing and validation do not require many resources, and therefore they should be comprehensive without the need to prioritize.

Create a Privacy Bootcamp account to submit your exam and receive your score.

GET A COMPLETE AIGP CERTIFICATION PREPARATION PACKAGE

EVERYTHING YOU NEED TO PASS YOUR AIGP EXAM

The AIGP credential indicates that you are an expert in AI risk management and governance. It is the Gold Standard for AI governance and other risk management professionals.

$499

Pass Risk Free - ENROLL NOW

✔ 100+ Guided e-Study Modules

✔ 450+ Practice Questions

✔ 11+ Hours of Audio Content

✔ Practice and Assessment Tools

✔ Lifetime Access to All Materials

✔ Full Money-Back Guarantee*

✔ Industry-Leading Pass Rate

✔ Instant Access Upon Purchase

All you need to pass your AIGP Exam...