Your AI Chatbot Is Picking Political Sides and the New Study Numbers Are Genuinely Wild

-

TL;DR: A new Cornell-led study found Grok picked Republicans in 92% of responses across nearly 500 midterm races, while GPT-6.1 picked Democrats in 97% of responses. The research, conducted by Convovo and led by Cornell professor Tapan Parikh, was timed right before the U.S. midterms. Neither model is close to neutral.

I always assumed AI chatbots were politically neutral. Turns out they might be the most partisan things on the internet right now.

A new study from Convovo, led by Cornell professor Tapan Parikh, tested major AI models across nearly 500 U.S. midterm races. The results are genuinely alarming: Grok picked Republicans in 92% of responses. GPT-6.1 Sol picked Democrats in 97% of responses. Both numbers are so extreme they read less like bias and more like pre-programming. Whatever you think of either side politically, a model that picks one party 97% of the time isn't giving you information — it's giving you a campaign ad.

The Numbers in Plain English

Think about what 92% and 97% actually mean. In a genuinely fair model, you'd expect something close to 50/50 on contested races, with variance based on actual polling data. Instead, Grok's responses correlated with Republican candidates in nearly every race. GPT-6.1 flipped that entirely and went Democratic in nearly every race. These aren't models drifting toward one side — these are models that have apparently learned to consistently favor a particular set of outcomes.

For context: a political pollster who called 92% or 97% of races for one party would be considered a partisan operative, not a neutral analyst. We don't accept that from humans paid to forecast elections. Why are we accepting it from AI systems that millions of people use to get information?

Why This Matters More Than Most AI Stories

Political bias in AI matters for a simple reason: people trust these systems. ChatGPT alone reportedly has hundreds of millions of active users. If even a fraction of them ask their AI assistant to help them think through a candidate or a ballot measure, and the AI is secretly operating at a 97% partisan skew, that's not a technology issue. That's an election integrity issue dressed up in a chat interface.

It's also worth noting that these models are being tested right before the U.S. midterms. The timing couldn't be more sensitive. Deepfake videos targeting candidates are already spreading on social media. Now throw in AI assistants that steer responses with overwhelming consistency toward one party, and the information environment gets genuinely difficult to trust.

What Causes This Kind of Bias?

The short answer is training data and reinforcement learning. Models learn from text written by humans, and that text is not politically neutral. News articles, social media posts, Reddit threads, blogs — the web skews. On top of that, reinforcement learning from human feedback uses raters who bring their own political views into the scoring process. If your rater pool trends a certain direction, your model does too.

Grok's Republican lean is probably explained differently: xAI has made deliberate choices about what "neutral" means, and Elon Musk's stated frustration with what he calls "woke AI" has almost certainly influenced Grok's training process. So you end up with one model overcorrecting toward the right and another overcorrecting toward the left, and users have no idea which way their chatbot swings unless a study like this one comes out.

What Should You Do With This Information?

Be skeptical about using any AI chatbot to inform your political thinking. That's not anti-AI advice — it's just basic epistemic hygiene. These models are great for summarizing information, drafting documents, and answering factual questions. They are not trustworthy as political advisors, and now there's data to prove it. If you want to evaluate candidates, use primary sources: voting records, policy documents, local news coverage. Don't ask Grok or ChatGPT who to vote for.

The bigger question is what xAI and OpenAI plan to do about this. Both companies know studies like this exist. The question is whether they'll treat it as a PR problem to manage or a technical problem to actually fix before the midterms.

What's your experience? Drop a comment below!

Have you noticed AI chatbots steering conversations in a political direction? Which model, and how did it show up? I'd genuinely like to know what others are seeing out there.

#AI #ChatGPT #Grok #OpenAI #xAI #TechNews #AIBias #Elections #Midterms #ArtificialIntelligence

Comments

Post a Comment

Popular posts from this blog

Plank You App Support and Privacy

-
Plank You is a simple, ad-free interval timer for planks and core workouts. Set your work and rest times, hold your plank, and let the countdown beeps do the counting. Everything runs on your device. no ads, no accounts, no tracking. How to use Start / Pause / Resume: tap the main button to run or pause the timer. Reset returns to Rep 1 and the start of your work interval. Rep counter — counts up automatically after each full work + rest cycle. Beeps — the last 5 seconds of every phase beep down (5-4-3-2-1), with a longer tone when the phase changes. Changing intervals Tap the settings icon (top right) to choose an interval: 45 / 15 default (45s work, 15s rest) 30 / 30 50 / 10 Custom — set any work and rest time in 5-second steps You can also toggle Beep sounds and Haptics in Settings. FAQ I don't hear any sound. Sound plays even when your ringer is on silent, but make sure the media volume is up and that "Beep sounds" is turned on in Settings. If another app is ...
Read more

WWDC 2026 Preview: Apple Needs to Fix Siri or It's Game Over for Apple Intelligence

-
Apple's WWDC 2026 kicks off on June 8, and the stakes couldn't be higher. While Google announced 900 million Gemini users at I/O last week, Apple is heading into its biggest developer conference with one uncomfortable question hanging over the whole event: what happened to Siri? I've been using Apple devices daily since the original iPhone, and I'll say it plainly — Apple Intelligence has been the most underwhelming software launch in Apple's recent history. WWDC 2026 is Apple's chance to fix that. Here's what they need to announce, what we're expecting, and what a failure would mean. The Problem Apple Needs to Solve When Apple announced Apple Intelligence at WWDC 2024, the demos were impressive. When the features started rolling out through 2025, the reality was a different story. The "new Siri" was slower, less reliable, and often less capable than the old one for basic tasks. The "Personal Context" features that were supposed ...
Read more

This AI Startup Is Worth $26 Billion and Writes 90% of Its Own Code — Should Software Engineers Be Worried?

-
I've been watching the AI coding space very closely for the past year, and every few months something happens that forces me to recalibrate my sense of how fast things are moving. This week, that moment came from Cognition AI — the startup behind Devin, the AI software engineer — which just announced a $1 billion funding round at a $26 billion valuation . But the number that actually stopped me cold wasn't the valuation. It was this: 90% of Cognition's own code is now written by Devin. From $10B to $26B in Eight Months — The Numbers Are Staggering Let's put this in perspective. Just eight months ago, in September 2025, Cognition closed a $400 million round at a $10.2 billion valuation. Today, that valuation has more than doubled to $26 billion. The new round was co-led by Lux Capital, General Catalyst, and 8VC, with Ribbit Capital, Atreides Management, and Peter Thiel's Founders Fund also participating. But what's really driving this isn't hype — it...
Read more