This Week in AI Writing: When the Arms Race Hits Home
Week of December 8th, 2025
The big AI companies are in an all-out sprint—OpenAI just dropped GPT-5.2 this week after an internal “code red” memo about losing ground to Google’s Gemini 3.
Meanwhile, writers are watching this tech showdown and asking a different question: what happens to authenticity when the machines get this good?
That tension showed up everywhere this week.
Google partnered with major publishers to test AI-powered article overviews in Google News.
The Authors Guild’s “Human Authored” certification program got new attention as writers weigh how much AI assistance they can use before they’ve crossed a line. And underneath it all,
a growing conversation about whether the flaws in our writing—the imperfect, human bits—are actually what make it worth reading.
If you’re trying to figure out where you stand as AI tools get more capable, you’re not alone.
What I’ve Been Writing About
Over on my Substack this week, I published
This one’s about how writers who use AI effectively can accelerate and improve their craft. Those who use it as a ‘print button’ will fall behind.
I also wrote
about my experience using a new AI tool on my Working Desk fantasy trilogy. The tool is an AI persona I created that examines stories, outlines, or ideas and evaluates their marketability. I didn’t get great results! The tool is available to paid subscribers:
In addition, I’m working on a 90-minute course and could use your help. Could you take 2 minutes to fill out a 1-question (technically 4) survey? It would mean a lot.
What the Industry is Talking About
The AI Arms Race Goes Nuclear
OpenAI launches GPT-5.2 after “code red” scramble
Just a month after their last update, OpenAI released GPT-5.2 on December 11, positioning it as their “best model yet for everyday professional use.” The timing wasn’t accidental—CEO Sam Altman sent an internal memo raising alarms about ChatGPT losing consumer market share to Google’s Gemini 3, which has been dominating benchmarks (except in coding, where Claude Opus 4.5 still leads).
Why writers should care: These tools are evolving faster than we can keep up. Each upgrade makes them better at mimicking human writing—which raises the stakes for those of us trying to maintain authentic voice.
My take: The companies are competing on benchmarks and features, but they’re not competing on the thing that actually matters to writers: helping us write better while staying ourselves. That’s still on us to figure out.
Publishers Embrace—and Resist—AI
Google tests AI-powered article overviews on Google News
Google announced a commercial partnership program this week with major publishers, including The Guardian, The Washington Post, and Der Spiegel. They’re experimenting with AI-generated article summaries to give readers more context before they click through.
The New York Times and Chicago Tribune sue Perplexity AI.
Meanwhile, major news organizations are suing AI companies for copyright infringement in an ongoing wave of legal battles over the use of training data.
Why writers should care: Publishers are caught between wanting to experiment with AI and protecting their intellectual property. The lawsuits will shape whether AI companies need to license content or can train on anything they scrape.
My take: You can’t have it both ways. Either AI companies respect copyright and compensate creators, or they don’t. Google’s partnerships suggest one path forward—but litigation will determine whether it’s voluntary or mandatory.
The Authenticity Question Gets an Answer (Sort Of)
Authors Guild’s “Human Authored” certification gains traction
The Human Authored certification program launched in January 2025, but it’s getting renewed attention as AI tools get more capable. Writers and publishers can use it to indicate that the text of a book was human-written (you can still qualify if you use AI for spell-checking or research, as long as “the literary expression itself” came from human intellect).
New research on how writers actually use AI
A 2025 Conference on Creativity and Cognition study found that creative writers are intentional about integrating AI, making deliberate decisions guided by core values such as authenticity and craftsmanship. Some writers avoid AI even in areas where they struggle because of voice concerns.
Why writers should care: Transparency matters to readers. 74% of authors who use AI choose not to disclose it, but readers are increasingly asking for that transparency.
My take: I’m glad the Authors Guild created a certification, but I’m not sure it solves the real problem. The question isn’t whether AI was involved—it’s whether the work has a distinctive voice and something true to say. A human can write soulless content just fine without AI help.
Time Magazine Names “Architects of AI” Person of the Year
Time recognizes AI’s seismic impact
On December 11, Time announced that it has named the “Architects of AI” as its 2025 Person of the Year, acknowledging the technology’s transformative influence across industries.
Why writers should care: AI isn’t going anywhere. The question is how we adapt.
My take: Recognition is one thing. Understanding the ethical implications for creative work is another. I’m less interested in celebrating the architects than in holding them accountable for how their tools affect writers’ livelihoods.
Worth Reading This Week
AI: The Rise or Fall of Creative Writing?
Duke University’s English Department weighs in with a measured take: AI may help with editing, but its use must be limited to preserve the writer’s voice. The key insight: “Authentic writing is by nature imperfect, and it is these flaws that make a piece more enticing.”
From Pen to Prompt: How Creative Writers Integrate AI
This academic study (full paper available) interviewed 18 creative writers who regularly use AI. The findings: writers base their engagement with AI on core values like authenticity and craftsmanship, deliberately choosing when, why, and how to use it.
Best AI Writing Tools for Authors in 2026
Kindlepreneur’s comprehensive guide to the current landscape, updated for late 2025. Useful if you’re trying to figure out which tools might actually help your process.
The Publishing Workshops Taking a Red Pen to AI
Publishers Weekly reports that 63% of publishers are using AI in some way, but most aren’t enthused about it. Some editors are using “the decision not to use AI as something to set themselves apart,” believing that a human touch is worth protecting.
AI Licensing for Authors: Who Owns the Rights?
The Authors Guild breaks down the thorny question of compensation. Key stat: 90% of authors surveyed believe they should be paid if their books are used to train AI systems, and 65% back a collective licensing system.
Looking Ahead
The arms race between OpenAI, Google, and Anthropic will continue—expect more model releases, more capability announcements, more benchmark wars. But the more interesting question is how writers will respond.
We’re past the point of wondering whether AI will change creative writing. It already has. The question now is whether we can use these tools without losing what makes our work ours.
That’s the conversation I want to keep having. If you’ve been wrestling with where to draw the line on AI assistance, or if you’ve found a workflow that preserves your voice while leveraging the tools, I’d love to hear about it in the comments.
Housekeeping
If you’re getting value from these weekly roundups, consider upgrading to a paid subscription. It helps me keep writing and researching the intersection of craft and technology.
And if you’re new here, welcome. This newsletter is about maintaining an authentic voice in an age of increasingly capable AI. No hype, no panic—just thoughtful exploration of what it means to write well when the machines can write too.
Sources for this week’s news:
You did provide me some insights on how to combine my voice and craft with AI's help. Thanks