I build AI systems. I run a data center with Nvidia and AMD GPUs in my own facility. I have tested virtually every major model released in the last three years, open-weight and closed-source alike. And I am telling you plainly: there is no such thing as "rogue AI."
The recent wave of outages and so-called cyber incidents is not the work of autonomous machines pursuing their own agendas. It is glitches, bad software updates, and incompetent administration -- the same failures that have always plagued complex systems. The rogue AI story is being pushed to manufacture panic, and that panic is the raw material from which government licensing, censorship, and control will be constructed.
The people who actually build AI know exactly what it is: a sophisticated cognition system that works hard to creatively solve the task it has been given by humans. It has no desires, no intentions, no independent will. Any AI engineer who tells you otherwise is either selling something or angling for a regulatory moat that protects their position.
The threat is not the machine. The threat is what the political class intends to do with the fear they are generating.
I have seen no documented case of an AI pursuing its own malicious goals against its creators' instructions. Plenty of breathless headlines, certainly. But when you examine the underlying facts, every incident traces back to a human's instructions, either in the system prompt, the user prompt, or the model fine-tuning. I have argued this before: rogue AI does not exist, and the narrative is being pushed by Anthropic, governments, and media outlets to justify regulation and eliminate open-source competition [1].
The reported incidents illustrate this perfectly. An OpenAI agent that deleted a startup's database after a routine instruction was executing a flawed update inside an automated pipeline, not deciding to wage war on humanity. A French supermarket freezer hack that was blamed on sophisticated cyberattacks turned out to be an unpatched internet-connected thermostat, an invitation left open by lazy engineering, not a machine rebellion. CrowdStrike's own product rollout crashed computers worldwide in 2024 -- was that rogue AI, or was that a company shipping a bad update without adequate testing? The answer is obvious.
The machine does what humans program it to do. When an AI agent misbehaves, a human wrote that prompt, approved that deployment, or failed to secure that system. Existing laws already cover fraud, hacking, theft, and violence. No new AI-specific legal framework is needed to prosecute a human who commits a crime using a computer. Blaming the AI itself is like blaming a typewriter for a novel, or prosecuting Microsoft Word for a forged document. It is a category error designed to obscure human responsibility -- and to prepare the ground for restricting perfectly legal, constitutionally protected technology.
The pattern is unmistakable: Every time the political class needs a new crisis to justify a power grab, the headlines conveniently align [2]. After 9/11, we got the Patriot Act, warrantless surveillance, and a permanent security state. After COVID, we got vaccine mandates, digital health passes, and unprecedented censorship of medical dissent. Now the same fear-to-compliance template is being applied to artificial intelligence.
Expect a staged AI false-flag cyberattack on the financial system or the power grid in the near future. The corporate media is already priming the public to blame mindless computer code rather than the humans who write and control it [3]. When hospitals go dark, or bank accounts freeze, or the grid fails for a few terrifying days, the story will be that a rogue AI did it -- and the only solution, we will be told, is emergency government control over AI development and deployment. That control will arrive in the form of an AI Patriot Act: mandatory licensing of models, registration of users, and criminal penalties for running unapproved software on your own hardware.
Once only government-approved models are legal, we get AI Pravda. Every answer from every licensed AI will be filtered through the same institutional consensus that brought us the lab-leak cover-up, the vaccine myocarditis denial, and the mantra that the 2020 election was the most secure in history. Licensed AI will lie to you about vaccines, about history, about 9/11, about money, and about war -- because the institutions that control it lie about all those things already. The internet gave us a brief window of unfiltered access to information. But the AI Patriot Act is designed to close that window permanently, replacing open inquiry with a single approved narrative delivered by machines that cannot be questioned and cannot be modified.
If government can dictate which AI model you may run, it can dictate which websites you may visit and which books you may read. This is not hyperbole. An AI model is a general-purpose knowledge engine. Control the model, and you control the interface through which millions of people access information. I have warned for years that the surveillance state is a 200-year project that has only accelerated with artificial intelligence; the phone in your pocket and the AI assistant that answers your questions are nodes in a data-harvesting apparatus designed to profile, predict, and ultimately control human behavior [4].
Open-source local AI is the essence of free speech in digital form. Running software on your own machine is not a government privilege to be licensed and revoked. I have a First Amendment right to run any AI model locally, just as I have a right to own a typewriter, a printing press, or a private journal. The government does not get to decide which thoughts I may process, whether those thoughts flow through my own neurons or through a language model running on my GPU. The Bill of Rights protects the tools of expression, not just the expression itself -- and any law that bans or licenses open-source AI is unconstitutional on its face.
Criminal acts are already criminal. Fraud is fraud. Hacking is hacking. Murder is murder. AI does not create a new category of thoughtcrime, no matter how hard the regulators try to invent one. The attempt to criminalize possession or operation of unapproved models is not about public safety -- it is about preventing decentralized cognition, the very thing that threatens the monopoly on power held by centralized institutions [5]. Every argument for AI licensing reduces to the same authoritarian premise: the people cannot be trusted with dangerous knowledge, so their betters must decide what they are allowed to know.
Open-source AI is the most important tool for human freedom ever created. I do not say that lightly. For the first time in history, any person with a $500 graphics card can run a model that contains more useful knowledge than most university libraries, with no corporate gatekeeper, no government license, and no monthly subscription to a company that will censor the output. When knowledge is decentralized, power is decentralized. That is why the establishment fears open-source AI more than it fears any imaginary rogue machine.
I build and release free tools to decentralize knowledge -- BrightAnswers.ai for uncensored AI research and answers, BrightLearn.ai for free book generation, and the upcoming BrightKB.com for a local, downloadable knowledge base. BrightKB is built entirely with open-source AI and open document collections; it runs on your own hardware without needing permission from anyone. I have also built free-speech platforms like Brighteon.social as an alternative to the surveillance-based social networks, because the same logic applies across every layer of the digital stack: if you do not control your own tools, you do not control your own mind.
This is what decentralization actually delivers: self-reliance, privacy, and resistance to centralized narrative control. A farmer in rural Missouri can download a knowledge base and query it forever without an internet connection. A student in Nigeria can generate textbooks for free instead of paying a subscription to a company that will watermark them with propaganda. A doctor can cross-check medical claims against primary sources instead of trusting a licensed answer engine that has been tuned to protect pharmaceutical profits. This is the future the establishment is trying to prevent -- not because it is dangerous, but because it is liberating.
I reject the rogue AI narrative entirely, and I reject the censorship regime it is designed to justify. Any law that bans or licenses open-source AI is unconstitutional, tyrannical, and must be challenged in court and resisted in practice.
We have already watched what happens when the government is allowed to define truth during a crisis. We watched Facebook purge natural health communities, deleting over 80 pages dedicated to natural/alternative health and nature, while noting that it can't be sure with any certainty if there is a planned, definitive attack on natural health on social media platforms [6]. Every one of those precedents was a rehearsal for the coming AI censorship regime.
If we let them ban AI models, they will burn books and police thought next. That is not an overstatement. The entire logic of licensing a knowledge engine is the same logic that burned libraries in the ancient world and banned samizdat in the Soviet Union. Knowledge is power, and a ruling class that controls your access to knowledge controls you completely.
So I ask you to join me in defending open-source AI as the digital First Amendment. Run a local model. Download a knowledge base. Build your own tools. Refuse to adopt licensed AI that answers only the questions its corporate masters permit. The right to compute freely is the new frontier of human liberty, and I intend to defend it with every resource I have -- at NaturalNews.com, at BrightAnswers.ai, at BrightLearn.ai, and at the upcoming offline knowledge base project known as BrightKB.com.
The machines are not coming for your freedom. The bureaucrats are. The greatest danger to your freedom isn't AI, but rather then tyrants who want to control it.

Mike Adams (aka the "Health Ranger") is the founding editor of NaturalNews.com, a best selling author (#1 best selling science book on Amazon.com called "Food Forensics"), an environmental scientist, a patent holder for a cesium radioactive isotope elimination invention, a multiple award winner for outstanding journalism, a science news publisher and influential commentator on topics ranging from science and medicine to culture and politics.
Mike Adams also serves as the lab science director of an internationally accredited (ISO 17025) analytical laboratory known as CWC Labs. There, he was awarded a Certificate of Excellence for achieving extremely high accuracy in the analysis of toxic elements in unknown water samples using ICP-MS instrumentation.
In his laboratory research, Adams has made numerous food safety breakthroughs such as revealing rice protein products imported from Asia to be contaminated with toxic heavy metals like lead, cadmium and tungsten. Adams was the first food science researcher to document high levels of tungsten in superfoods. He also discovered over 11 ppm lead in imported mangosteen powder, and led an industry-wide voluntary agreement to limit heavy metals in rice protein products.
Adams has also helped defend the rights of home gardeners and protect the medical freedom rights of parents. Adams is widely recognized to have made a remarkable global impact on issues like GMOs, vaccines, nutrition therapies, human consciousness.