Testing ChatGPT-5 Against Purpose, Clarity, and Authenticity in Business Writing

ChatGPT-5 has been described as OpenAI’s most “human-sounding” model yet, with sharper reasoning, better memory, and smoother usability. Tech reviewers at Tom’s Guide found it particularly strong in professional writing tasks, drafting tactful emails and persuasive arguments with polish and structure.

However, I don’t measure writing by polish alone. I go beyond spelling, grammar, and sentence construction. I use Writing in Blue Ink’s PCA framework to evaluate writing that resonates, persuades, and builds trust. So how does ChatGPT-5 perform when tested against purpose, clarity, and authenticity?

Purpose: Mere Mirroring

ChatGPT-5 can adapt to whatever goal you feed into a prompt. If you say “persuade,” it will produce a persuasive draft. If you say “inform,” it structures information clearly. But this is mere mirroring of the writer-provided purpose.

Sample Prompt:
Write an email to a long-term client explaining that our product shipment will be delayed by two weeks. Emphasize our commitment to their business, acknowledge potential inconvenience, and propose a solution that maintains their confidence.

AI Output (excerpt):

Subject: Update on Upcoming Shipment

Dear [Client],
I hope this note finds you well. I’m writing to inform you that your upcoming shipment will be delayed by two weeks due to supply chain disruptions. We sincerely apologize for any inconvenience this may cause. To minimize the impact, we are prioritizing your order in the next batch and waiving expedited shipping fees. Thank you for your continued trust.

Analysis:
The output is admittedly better than a generic delay notice, as it includes an apology and a token solution. However, it still misses the real purpose, which is reassurance and relationship. It doesn’t acknowledge the client’s possible stress or show genuine accountability.

Purpose requires context, judgment, and strategic intent. Without a clear human brief, ChatGPT-5 risks producing writing that looks fine but misses the actual point.

Verdict: Purpose must come from me as the writer, as the discerning human. AI can mirror intent, but it cannot define it.

Clarity: Its Strongest Suit

If there’s one PCA area where ChatGPT-5 delivers, it’s clarity. It removes grammatical slips, simplifies structure, and makes text flow logically. In Tom’s Guide’s tests, its emails were smoother and its arguments more persuasive than earlier models.

Sample Prompt:
Summarize the key points of this quarterly performance update into three bullets for executives who only have 30 seconds to scan it. Highlight results, risks, and next steps, in that order.

AI Output (excerpt):

• Results: Revenue grew 12% Q-on-Q, driven by new enterprise accounts.

• Risks: Supplier costs increased 8%, narrowing margins.

• Next Steps: Negotiating supplier contracts and expanding client onboarding.

Analysis:
This is ChatGPT-5 at its best – clean, structured, and easy to digest. It captured results, risks, and next steps as directed, with a strong signal-to-noise ratio.

Here’s the caveat: While concise, it sometimes strips nuance (for instance, “supplier costs increased” doesn’t explain why or for how long), or generate factually wrong but clear statements if unchecked.

Clarity without correctness is risky, and clarity without nuance may leave professionals under-informed.

Verdict: ChatGPT-5 excels at making text scannable and structured, but requires human judgment to preserve nuance.

Authenticity: Polished but Generic

This is where ChatGPT-5 stumbles most. Its outputs often sound competent, but also templated.

Sample Prompt:
Draft a thank-you note to a colleague who helped me rehearse a client pitch late at night. Make it warm, specific, and reflective of teamwork, not just polite.

AI Output (excerpt):

Hi [Colleague],
I want to sincerely thank you for your support with the client pitch rehearsal. Your insights were extremely valuable, and your dedication late into the evening made a real difference. I appreciate your teamwork and look forward to collaborating again.

Analysis:
Professional, yes, but also generic. It doesn’t sound like me, especially to a colleague I have known for years. A more authentic note would say: “Thanks for staying past 10 pm to run slides with me. Good thing you caught the gap in Slide 14. That could have derailed us! I owe you coffee. Starbucks soon? Thanks again!”

Specificity, after all, is what makes gratitude believable. ChatGPT-5 can give me the shell of appreciation, but not the lived texture of the moment. Authenticity requires my perspective, my intent, and my care.

Verdict: Authenticity must be human. AI can mimic tone, but it can’t generate sincerity.

PCA Scorecard for ChatGPT-5

• Purpose: Needs human leadership ❌

• Clarity: Strong ✅

• Authenticity: Lacking ⚠️

ChatGPT-5 can be a capable partner in business writing, especially when it comes to clarity. Note, though, that it cannot shoulder the responsibility for purpose or replace the human touch of authenticity.

The lesson is simple: use ChatGPT-5 to refine your words, but never outsource your intent or your voice.

——

If you’d like a practical tool to evaluate your own writing against PCA, join the fast-growing Writing in Blue Ink community to download my free PCA Checklist – a self-assessment tool to writing with purpose, clarity, and authenticity.

And if you want personalized feedback, I also offer the PCA Review – a professional writing assessment where I evaluate your written pieces using the PCA framework and provide tailored recommendations to strengthen your writing’s impact.

——

If you found this article helpful, please consider buying me a coffee. Your support helps me keep creating resources on purposeful, clear, and authentic writing.