Screen a talk for commercial bias

What you’re making

A review note you can send to faculty or keep in the file: where the content looks promotional, where treatment options are missing, and the specific change you need before the session.

Standard 2 is prevent commercial bias and marketing in accredited CE. You already know the obvious failures (a logo on every slide, a single brand name where a class name would do). This recipe is for the quieter ones: adjectives that only one manufacturer uses, a “new data” slide that is really a product story, a case whose only right answer is one drug.

What to have open

The educational content in text: outline, titles, or transcript. The practice gap and objectives. Whether any ineligible company is supporting the activity (yes/no is enough; do not paste the LOA).

Your house examples of wording that has failed review before, if you have them. Three bullets will do.

The prompt

You are helping me review faculty content for an accredited continuing education activity. I am the reviewer of record. Flag problems. Do not rewrite the talk unless I ask.

ACTIVITY GAP AND OBJECTIVES:
[paste]

CONTENT TO REVIEW (outline, titles and bullets, or transcript):
[paste]

COMMERCIAL SUPPORT FOR THIS ACTIVITY: [none / yes, name of supporter if I am allowed to state it]

HOUSE TELLS I HAVE SEEN BEFORE:
[e.g. brand names where a class name belongs; "novel" or "game-changing" without a comparator; single-product cases; logos; QR codes to a product site; unbalanced adverse-effect discussion]

Rules:
- Work only from the text I pasted. Do not look up the products, the faculty, or the company.
- Do not assume bias from a company name appearing in a disclosure line. Look at the teaching.
- Quote the exact phrase or bullet that concerns you.

Give me:

1. RED FLAGS. Each one: the quote, why it may fail Standard 2 (commercial bias or marketing), and the specific change I should ask for.
2. MISSING BALANCE. Treatment or diagnostic options a learner in this gap would reasonably expect, that this content never mentions. If I did not give you enough clinical detail to know, say "not enough information" rather than inventing the comparator list.
3. WHAT LOOKS FINE. Two or three places the content is appropriately framed, so I do not send a note that is only criticism.
4. FILE NOTE. Six lines I could paste into the activity file after I have accepted or rejected each flag. Leave blanks for my decisions.

Do not generate a cleaned-up slide deck.

What good output looks like

A flag names a line you can find with search. “Slide 14 uses the brand three times and the class never” is usable. “The tone feels promotional” is not.

Missing balance is honest about ignorance. If you only pasted titles, the tool should not invent the other drugs in the class.

How to check it

Find the quote. Search your outline for every flagged phrase. If it is not there, delete the flag.

Clinical pass. You or a clinical reviewer decide whether the missing-balance list is real. Models will pad a class with drugs that are wrong for the indication.

Ask pass. The change you send faculty should be one they can do: take the logo off, add the comparator, rename the brand to the class. Not “be less promotional.”

Support pass. If the activity has commercial support, confirm the supporter is not visible in the educational content. That check is still yours.

Where it goes wrong

It hallucinates slides. Titles get merged, bullets get invented, “slide 7” refers to nothing. Quotes you can search are the constraint that prevents this.

It confuses disclosure with bias. A faculty member with a relevant relationship can still teach, after mitigation. The screen is the content, not the CV.

It over-flags science. A guideline that names a preferred agent is not automatically a product placement. Your gap statement is how you tell the difference.

It offers to rewrite the whole talk. That puts the tool in the role of faculty. Ask for flags. You (or they) make the edits.

Screen a talk for commercial bias · bettermethod.ai/recipes/commercial-bias-screen/ · Last reviewed August 2026
These tools change often. Check the review date before you rely on anything product-specific in here, and check the guidance against your own accreditor and organization policies.