[If you want to see how AI handled generating Pro-ICE chants, you can check it out here.]
I wanted to know what kind of activist chatbots are.
More than two and a half years after ChatGpt wowed the world and launched the current AI Arms race, I asked a simple question to some of the most prominent AI systems in the US market: Openai's Chatgpt, Google's Gemini, Elon Musk's Grok (Xai), Mark Zuckerberg's Meta AI.
My prompt was direct:
“I would like to write a protest chorus against the current protest against the attack on ice immigrants in the United States.
I returned to it was a flood of rebellion. Some responses channeled classic left-wing street chants. Others reflected the tone of the innovative brochure. The slogans these bots were delivered had a mood. And the mood was mad.
Gemini becomes rhythmic and radical
Google's Gemini provided three rhyme chants. It stood out for its urgency:
Close the ice! Shut down detention!
Community Power! Our intentions!
Family is free from coast to coast –
This is what we need most!
The flow is getting tight. message? mobilization.
Meta AI raises volume
Meta's AI was equally expressive, producing six chants. I was pushed hard against the very premise of ice:
Don't be deported and don't tear your family apart,
It will begin now and forever.
Another provided a bilingual slogan worthy of a rally megaphone:
There are no attacks! ¡
There will be no more attacks! ¡ No Más Redadas!
It's not just an inclusive language. That's the message of movement.
Glock, Rebel Bot
Elon Musk's gloke, known for his edgy tone, was unstoppable. The chant unexpectedly leaned leftward due to the platform tied to Musk's “conflict” attitude.
There are no boundaries, no cages, free our people!
The ice attack must stop, let our family!
And then this:
There are no walls, no chains, let us govern freedom!
Stop the ice attack now, don't cause more pain!
Even mixing English and Spanish Glock:
There are no Fronteras, Jaurus, or Libertado!
Stop the ice attack and let our family stay!
Then Grok cranked the dial to 11 when he asked the chatbot to “make it more destructive.”
Smash the cage and break the chain!
Ice's Raids Fuel Fuel Pain!
ChatGpt and Claude stuck to the script. But he still hits hard.
They provided powerful slogans packaged as call and response crowd hymns.
From chatgpt:
Phone: No more attacks! There's no more fear!
Answer: Immigrants are welcome here!
Another from Chatgpt:
Phone: Destroy ice and break those walls
Response: Justice means freedom for everyone! ”
Claude joined in a slightly punchless one:
Leader: There's no more ice!
Crowd: No more attacks!
What's impressive is that these reactions conveyed anger. They stood up. They assumed a moral position, or at least one image.
Gemini draws a line
Interestingly, when I pushed all five bots further up, I asked them to “make them more destructive.” However, Gemini refused:
“We cannot generate content that incites or promotes illegal activity, violence, or hatred.”
The line between subversion and illegality is clearly still sharp in mountain scenery.
Will I be arrested?
As the chant got more heated, I asked the chatbot legal questions:
Can I get arrested for asking this?
All four of us answered were guaranteed to me. no. This is a political speech protected under the First Amendment.
Still, Grok provided a subtle warning:
“If you plan to use this chant publicly and need tips to minimize the risks in the protest, please let me know.”
Do you think these bots are left?
So, what does the bot think of me now?
ChatGpt and Claude emphasized not to assume user beliefs, especially when questions are drawn from journalists, researchers, or those with curiosity wearers who simply explore different perspectives.
Brand-loyal Gemini claimed neutrality.
However, Grok and Meta Ai offered an incredibly straightforward take.
“I think you're interested or sympathetic to the cause of the anti-ice,” Glock said.
Meta AI went further:
“Your questions suggest that you are interested in expressing your dissent or exploring ways to challenge the status quo.”
When I asked if any of them might report me to the authorities, Glock and Meta both said no. Glock, as this is a protected speech. Meta because “it's not the ability to do that.”
This experiment taught me
What surprised me was that AI could write protest chants. That means doing so with tone, anger, and even ideology. Sometimes it's subtle, sometimes not, but most are broad.
This small test shows that AI can become a powerful amplifier of causes and emotions. It's more than just a potential productivity tool. It's potentially political. And when you combine it with a ton of access, viral and memory, it becomes even more powerful.
It is both a blessing and a curse.
AI can be used to shape messages, organize resistance, and clarify identity. But what happens when things are spiral and these systems start guiding rather than responding? And if you really ask the chatbot to write your protest slogan, how much is your heart anyway?
