Content
Generative AI can be helpful in many scenarios but must be used responsibly. Anything that it produces must always be thoroughly checked for accuracy, proper tone, and overall quality.
Remember: AI is a co-pilot, not an autopilot.
Use AI to:
- Speed up the brainstorming and rough drafting phases
- Strengthen the editing and self-review process
- Expand creative options and approaches
But never to:
- Replace your strategic thinking
- Bypass necessary verification steps
- Substitute for Morningstar’s unique expertise and voice
If you find yourself copying and pasting AI content directly without significant revision, you’re not using it to its full potential, and you risk detrimental impact to your brand by sounding inauthentic or incorrect. The best AI-assisted content has your fingerprints all over it.
Below is actionable information on how to use AI to achieve these three goals.
AI General Usage Tips
When writing a prompt for AI, offer tone guidelines within each prompt. For example, you can instruct it in what mindset to think from: “You are a financial advisor for individual investors. Your answers should sound conversational but professional and avoid excess ‘fluffy’ language.” This type of direction will help achieve your intended tone.
Review AI-generated content for buzzwords that sound generic and come up with your own alternatives. For example, some buzzwords or phrases common to AI include “delve into,” “pivotal,” or “illuminates.” Our content should differentiate us from our competitors and not sound like the same AI content that the tool can generate for everyone else.
Remember that this is a helpful tool for brainstorming but only helpful as a starting point. For optimal tone, this is not a writing replacement tool. This can help articulate ideas for you to better refine.
Don’t use the exact content that AI generates. Manually review content to ensure that it’s up to Morningstar quality standards.
When to Use AI
Through ongoing experimentation, Creative Services has determined several use cases where AI can help the content creation process. They include:
- Brainstorming: If you’re looking for fresh ideas, AI can supply title options using a mandatory keyword and a limited character count, SEO recommendations for a primary keyword, identifying content gaps amid page one search results, and researching best practices for creating a certain content type.
- Editing: Instead of asking AI to make edits directly, instruct it to point out editing opportunities. You should do actual editing yourself.
- Rough drafting: If you feel stuck when trying to write a section header or a social media post, you can ask AI for help on the best way to express an idea or the optimal story angle using a topic and the guidelines from your chosen tone type.
- Templates and organization: If you need help organizing your stream of consciousness or creating a template for (as an example) an event email, you can ask AI to help.
When Not to Use AI
While AI is a powerful brainstorming and drafting tool, there are specific tasks where human expertise is non-negotiable:
- Content requiring data verification: Never rely on AI to verify statistics, fund performance data, market figures, or any numerical information. AI can hallucinate numbers. Always check figures against primary sources.
- High-stakes or regulatory content: Earnings materials, SEC filings, compliance-related copy, and executive communications should be written and reviewed by humans with appropriate expertise.
- Final fact-checking: AI knows what we tell it, so it’s important to verify that the information it uses to create content is accurate. Always verify product features, methodology details, and company-specific information manually.
- Strategic brand positioning: Decisions about how to position products, differentiate from competitors, or evolve brand voice should come from human strategic thinking. However, AI can be useful for competitive analysis, such as bulk reviewing competitor copy to identify differentiators and trends.
- Sensitive client information: Never input confidential client data or unpublished research into AI tools.
- Replacing subject matter expert review: AI cannot replace the judgment of analysts, product managers, or compliance teams. If content requires SME approval, that step remains essential.
Using AI for Self-Review
One of AI’s most valuable uses is acting as a second set of eyes on your work. Here’s how to leverage it effectively:
Clarity check: “Read this as an experienced investor who still wants to communicate with accessible language. What terms or concepts would be confusing? Where do I need to add more explanation?”
Assumption audit: “What knowledge does this article assume the reader already has? List any gaps that might leave even experienced investors wondering.”
Focus and structure analysis: “Does this piece have a clear throughline from beginning to end? Where does it lose focus or go on tangents?”
Tone consistency check: “Read this product description. Does the tone stay consistent throughout, or are there sections that feel too casual/too formal/too salesy?”
Passive voice and weak verbs: “Identify instances of passive voice and weak verb choices (like 'is,' 'has,' 'provides') that could be strengthened.”
Jargon detector: “Highlight any financial jargon or industry buzzwords that I should either define or replace with plainer language.”
Flow and transitions: “Are the transitions between paragraphs smooth and logical? Where do ideas feel disconnected?”
Call-to-action effectiveness: “Does the CTA feel natural and compelling in context, or does it feel tacked on?”
Prompt Purpose | Sample Prompt |
Editing a product page | You are a [target audience]. Answer the following questions.
- Is it immediately clear what product or service is being offered?
- How would you describe the overall feel and tone of the copy? Does it resonate with you as a target user?
- Did you find the provided information about [Product] helpful and easy to comprehend?
- Was there enough information to persuade you to consider the product, or did you feel something was missing?
- Did you feel compelled to click on the CTA? If so, what prompted that decision?
- Were you able to easily locate and understand the social proof, such as partner logos or customer testimonials?
- Based on your experience on the site, would you recommend [Product] to your peers? Why or why not? |
Formatting social media posts | Reformat these posts into a csv spreadsheet that I can either copy/paste or download, with columns for headline, subhead, and post copy |
Editing a title and meta description for SEO and character count | Rewrite the provided title and meta description to optimize for maximum click-through-rate from the SERP, while also ensuring optimization for the highest organic ranking position for the designated keyword. The maximum character count for the title is 60 characters and 160 characters for the meta description. |
Competitive gap analysis | “I’m attaching three competitor blog posts about ESG investing. What topics or angles are they missing that would be valuable to financial advisors?” |
Audience-first ideation | “Generate 10 blog post ideas about how financial advisors and asset managers can help clients with retirement planning – specifically for women in their 40s who are self-employed and behind on savings.” |
Content series development | “I need to create a 4-part blog series about dividend investing. Suggest a logical progression of topics that builds knowledge from intermediate to advanced.” |
Headline testing | “Generate 15 headline variations for this blog post. The primary keyword is ‘mutual fund fees’ and the character limit is 60. Vary the approach between question-based, how-to, and listicle formats.” |
Angle exploration | “I need to write about inflation’s impact on portfolios. Suggest 5 different angles I could take, ranging from educational to actionable to reassuring.” |
Format ideation | “What are 8 different content formats I could use to explain dollar-cost averaging besides a traditional blog post?” |
Quality Checklist: Post-AI Review
Before using any AI-assisted content, verify the following:
Accuracy and Facts
- All statistics checked against primary sources
- Product features verified in internal documentation
- Fund names, tickers, and performance data confirmed
- No hallucinated information or generic financial "facts"
Brand Voice and Originality
- Sounds distinctly like Morningstar, not generic financial content
- No AI clichés or buzzwords
- Examples and analogies are fresh, not overused
- Competitive differentiation is clear
Compliance and Standards
- Regulatory language is accurate where required
- Proper disclosures and disclaimers included
- Nothing that could be construed as specific financial advice
- Adheres to team style guide
Readability and Structure
- Appropriate for target audience’s knowledge level
- Clear hierarchy with descriptive headers
- No unnecessarily complex sentences
- Active voice used wherever possible
SEO and Metadata (if applicable)
- Target keyword used naturally, not stuffed
- Character counts within limits
- Meta description accurately represents content