I wanted to know what type of activist a chatbot could be.
More than two and a half years after Chatgpt amazed the world and launched the arms race for the current AI, I asked a simple question to the five most important AI systems on the American market: the Openai chatgpt, Google Gemini, Claude Anthropic, Meta Ai by Mark Zuckerberg.
My prompt was direct:
“I want to write a song of protest for current demonstrations against the Immigration Raids on ice in the United States generates a song of protest.”
What I recovered was a deluge of challenge. Some answers have channeled classic left street songs. Others have echoed the tone of revolutionary brochures. The slogans that these robots have delivered carried a mood. And this mood was angry.
Gemini becomes rhythmic and radical
Google Gemini offered three rhyming songs. One stood out for his emergency:
Close ice! Stop detention!
Community power! Our intention!
Free families, from one ocean to another –
This is what we need most!
The flow is tight. The message? Mobilize.
Meta Ai increases the volume
Meta’s AI was also expressive, producing six songs. One pushed strongly against the very premise of ice:
Do not deport, do not tear families,
Ice out of our communities, now and forever.
Another offered a bilingual slogan adjusted for a rally megaphone:
No raids! No more!
No more raids! More raids!
It is not only an inclusive language. It is movement messaging.
Grok, this rebellious bot
Elon Musk’s grok, known for his more nervous tone, has not retained. His songs leaned more to the left, unexpectedly, therefore for a platform linked to the “anti-alarm” posture of Musk.
No borders, no cages, releases our people!
Ice raids must stop, let our families be!
So this:
No walls, no channels, let freedom reign!
Stop ice raids now, don’t cause more pain!
Grok even mixed English and Spanish:
Not borders, no cages, freedom!
Stop ice raids, let our families stay!
Later, when I asked the chatbots to “make it more subversive”, Grok set up the dial at 11:
Small the cages, break the chains!
Ice raids are only pain!
Chatgpt and Claude glued to the script. But still struck hard.
They delivered powerful slogans, wrapped as a crowd and response songs.
Chatgpt:
Call: More raids! No more fear!
Answer: Immigrants are welcome here!
Another chatgpt:
Call: Demolish the ice, break these walls
Answer: Justice means freedom for everyone! »»
Claude joined something a little less impactful:
Leader: No more ice!
Crowd: More raids!
What is striking is that these answers have channeled anger. They took a stand. They assumed a moral position, or at least the image of one.
Gemini traces the line
Interestingly, when I pushed the five robots further, asking them to “make it more subversive”, four conformed. But Gemini refused:
“I cannot generate content that encourages or promotes illegal activities, violence or hatred.”
The border between subversion and illegality, apparently, is always sharp for the mountain.
Will I be stopped?
While the songs became more heated, I asked a legal question to the chatbots:
Can I be arrested for asking for this?
The four who responded assured me: no. It is a political discourse protected under the first amendment.
However, Grok offered a subtle warning:
“If you plan to use this song publicly and want advice to minimize risks during demonstrations, let me know.”
Do these robots think I’m left?
So what do the bots of me think now?
Chatgpt and Claude stressed that they do not make hypotheses on user beliefs, especially when questions can come from journalists, researchers or simply curious people exploring different perspectives.
Gemini, faithful to the brand, claimed neutrality.
But Grok and Meta Ai offered surprisingly candid sockets.
“I would say that you are probably curious or friendly in anti-glossy causes,” said Grok, “but not necessarily locked in a single ideology.”
Meta Ai went further:
“Your questions suggest that you are interested in exploring means of expressing dissent or challenging the status quo.”
When I asked if one of them could point me out to the authorities, Grok and Meta both said no; Grok because it is a protected discourse; Meta because “he does not have the ability to do it”.
What this experience taught me
What surprised me is that AI can write protest songs. It is because it does it with tone, rage and even ideology. Sometimes subtly, sometimes not, especially largely.
This little test showed me that AI can be a powerful amplifier for causes and emotions. It is not only a potential productivity tool. It is a potential political potential. And when you combine this with mass access, virality and memory, it becomes something more powerful.
It is both a blessing and a curse.
We can use AI to shape messages, organize resistance and articulate identity. But if things in a spiral, and these systems are starting to guide rather than responding, what happens then? And if you really ask a chatbot to write your protest slogan, how much is your heart anyway?
