CBC News AI principles and practices

listen to this article

Estimated 9 minutes

The audio version of this article is generated by AI-based technology. Mispronunciations may occur. We work with our partners to continually review and improve our results.

To ensure full transparency, CBC News is sharing internal guidance for staff on the use of AI in reporting and journalism production. To learn more about CBC’s approach, please visit: CBC News editor’s blog.

1.0 Introduction

these principles and practices Establishing a framework for the responsible and ethical use of artificial intelligence (AI) at CBC News. Our core philosophy is AI has the potential to empower staff and enhance journalism. By carefully integrating AI tools, we aim to increase productivity and free up more time for demanding journalistic activities, while ultimately improving the viewer experience.

All CBC News employees are expected to respect these principles and practices, as well as the content they produce. These are rooted in the values outlined in journalistic standards and practices.

Because AI is a rapidly evolving field, these principles and practices will be continually reviewed, updated, and communicated. If in doubt, please contact your manager for guidance.

2.0 Core Principles and Goals

Our guiding principle is that CBC News employees, not AI tools, create and own the journalism.

2.1 Principles

Not the first draft: CBC News does not use AI tools in the initial stages of composing content intended to serve our viewers. AI-generated content should not be used in journalism unless there is an overriding editorial reason and requires prior approval from a manager or director.
Mandatory monitoring: AI is a tool. It is not the creator. Final editorial judgment, fact-checking and accountability rest with employees.
Accuracy and reliability: To maintain audience trust, all content, AI-assisted or not, must meet rigorous standards for verification and accuracy. This is achieved through a commitment to journalistic standards.
Transparency: We will be open with colleagues and audiences about how we use AI in our work.
Emphasis on skills: CBC News is committed to maintaining the skills and skills training of our journalists as a priority and preventing skill decline. At the same time, we will further expand AI literacy and training for journalists.

Use of AI is governed by: basic principles and strategic objectives that align with our goals. Journalism standards and practices.

All of the above principles apply to non-CBC News staff employed or contracted to assist in the production of this content while working for CBC News.

2.2 Goals

Improving employee experience: Automate and streamline routine tasks responsibly, freeing up journalism production staff for higher-value work.
Improve your audience’s experience: Use AI responsibly to, for example, deliver content in more accessible formats across different platforms.

3.0 Categories of AI use

It is important to distinguish between two major uses of AI. At CBC News, The use of AI focuses on AI support work:

4.0 Accountability and Responsibility

CBC News employees are responsible for all editorial content. This means:

We take responsibility for our results. The final job is ours. We are responsible for the accuracy, fairness, balance, impartiality and completeness of whatever tools we use.
Be able to explain your process. If you are asked these questions during the editorial review process, remember how you used the AI tool and be prepared to explain it.
Stay transparent: We will be open and honest with colleagues and audiences about how we leverage AI in our work. For legally sensitive stories, journalists must disclose their use of AI to their lawyers.

5.0 Permitted AI Assisted Functions and Practices

With the above core principles and goals in mind, AI tools have been approved for specific functionality to aid workflows. Our normal editorial and vetting processes apply to all scenarios.

Please note that there are some risks to using AI, including information security. It’s also important to be aware of automation bias and the tendency to trust or prioritize suggestions from automated systems even in the presence of contradictory information.

We also strive to be mindful of the environmental impact of using AI. These tools are best used in conjunction with prior CBC training on AI.

5.1 Permitted Functions

Approved uses fall into several categories.

Event planning/coverage: AI may be used to assist in the creation of planning documents for internal purposes.
Research and data analysis: AI could be used for research and data analysis, with an emphasis on fact-checking by journalists.
Summary and translation: AI may be used to create summaries and transcripts for internal use and to be checked and verified by journalists. The initial translation may also be done with AI, as long as the nuance and accuracy of the output is verified by a fluent speaker.
Story ideas: AI may be used to suggest story ideas. Be aware of who else is covering those stories and angles, and if their suggestions are not original ideas.
Reviews and feedback: AI may be used to improve, revise, and provide feedback on content created by employees and intended to be presented to audiences. When using AI-assisted content, you must respect the principle that journalists are the creators of the content. We are responsible for the accuracy, fairness, balance, impartiality and completeness of whatever tools are used. We openly share information and ask each other questions about the use of AI-assisted content, especially when it is incorporated into works for audiences. For legally sensitive stories, journalists must disclose their use of AI to their lawyers.

Notes on approved experiments: To innovate responsibly, the News AI Steering Committee may approve limited, short-term experiments to explore new AI tools and use cases. These projects have special supervision and are designed to support our learning. Participants will be clearly defined and the results will be used to inform future versions of these practices.

A note about content conversion: In certain approved instances where AI is used to transform content, such as text-to-speech or video subtitles, journalists are not required to be up-to-date before publication.

5.2 How to use News Team

employee Must use corporate Gemini and NotebookLM accountsIf you use AI for work, or an approved in-house CBC application.

Don’t use personal AI tools (e.g. ChatGPT, Claude) Prevent sensitive data from being leaked or content from being used as training data. The security level of information entered into Gemini and NotebookLM can be handled at the Google Docs and Gmail level. Highly sensitive materials are not recommended for input.

employee must Validate or cross-reference all AI outputwith an emphasis on fact-checking.

AI-generated content should not be used in journalism There is no definitive editorial reason to do so and prior approval from a manager or director is required. This includes, but is not limited to, image and video generators, the generated AI capabilities of our suite of tools, and generated AI content provided by our partners.

6.0 Audience Disclosure

Transparency is key to maintaining audience trust.

6.1 When to disclose

Disclosure is required if there is an editorial decision to include generated AI in content.

A general discussion of disclosure is strongly encouraged. If you need clarification on whether disclosure is required, please consult with your respective leadership team.

Examples where disclosure to viewers is required:

Analyzing huge datasets is impossible for journalists.
Automated text-to-speech.
Automatic subtitles.

Important questions to ask are: Is there a risk that your audience will be misled if you don’t disclose your use of AI? If the answer is yes, disclosure is required.

Please consider including the following details in your disclosure:

What did the AI tool do?
Why journalists used AI. Ideally, explain how AI has benefited and improved news reporting.
How journalists were or were not involved in the process and/or reviewed the content before publication.
A description of how the content meets the newsroom’s ethical and accuracy standards. Link to newsroom standards.
Examples of disclosures: In this article (AI/Tools/Tool Description) Please help us (What did the AI/tool do or help?). (if using)AI/Tools) we (Fact-checked and checked by journalists to ensure we meet journalistic standards). Using this, (Do more X, dig deeper, offer content on more platforms, etc.. ).

6.2 When disclosure is not necessary

There is no need to disclose the use of AI for routine support tasks that do not materially form the final editorial product. Examples include, but are not limited to: