Artificial intelligence (AI) companies are caught in a “perfect storm,” experts say.
The industry is trying to control the use and production of its platforms while maintaining profits and fighting campaigns to boycott the technology.
The QuitGPT protest movement against OpenAI’s ChatGPT has begun amid industry concerns that the Trump administration could use AI for autonomous warfare and mass surveillance.
Technology company OpenAI says its ChatGPT artificial intelligence platform is used by 900 million people around the world every week. (Reuters: Dado Lubitsch)
Sean Parnell, the Pentagon’s top spokesperson, posted on social media that the military “has no interest in using AI for mass surveillance of Americans.” [which is illegal] Nor do we intend to use AI to develop autonomous weapons that operate without human intervention. ”
Rival AI company Anthropic and the U.S. government are also embroiled in a legal cover-up over the Pentagon’s national security blacklisting of the company after the company refused to remove guardrails around the use of its technology.
The Trump administration has signed a deal with OpenAI to deploy its technology on sensitive networks, and the company has three “key red lines”: not to use the technology for domestic mass surveillance, not to direct autonomous weapons systems, and not to make “high-stakes automated decisions.”
The agreement intensified the QuitGPT movement, started by self-described “democracy activists” who were concerned that AI companies were contributing to “the rise of authoritarianism in the United States.”
What is the QuitGPT movement?
QuitGPT is a grassroots movement asking people to stop using ChatGPT, OpenAI’s underlying platform, and cancel their subscriptions, according to the movement’s website.
It claims to have attracted over 4 million participants. OpenAI says it serves more than 900 million users each week, including 50 million subscribers.
QuitGPT was launched in January amid a political storm caused by President Donald Trump’s deployment of Immigration and Customs Enforcement (ICE) agents across the United States.
The crackdown sparked protests in several cities, including Minneapolis, where federal agents shot and killed civilians Alex Preti and Renee Good.
More than 4 million people have joined an online protest aimed at boycotting the use of OpenAI’s ChapterGPT artificial intelligence platform. (Reuters: Manuel Orbegoso)
The QuitGPT movement points out that OpenAI president Greg Brockman donated US$25 million ($38.5 million) last December to the pro-Trump MAGA Inc. and Leading the Future super PAC, which was set up to promote “US leadership in AI innovation.”
QuitGPT’s website states, “Creating a ChatGPT example sends a clear signal to ICE enablers that their actions will go unpunished.”
U.S. law and Immigration and Customs Enforcement agents have clashed with protesters in multiple cities amid President Donald Trump’s massive deportation crackdown. (Reuters: Ryan Murphy)
The campaign also claims that users’ emotional dependence on AI platforms, the risk of people developing “AI psychosis” after “prolonged chatbot interactions,” the environmental impact of AI data centers, and the influx of advertising are reasons to boycott the technology.
An OpenAI spokesperson told ABC that Brockman’s political donations were made “in his personal capacity and not as OpenAI.”
The company “worked with more than 170 mental health professionals” to address emotional dependency concerns, a spokesperson said.
“While it’s right that people should care about how AI is developed and the companies that build it, this campaign makes misleading claims about OpenAI,” a spokesperson said in a statement.
“We are focused on building AI that is safe, convenient, and benefits as many people as possible.
“We will do that by making AI broadly available, investing heavily in safety, and working with governments around the world to help people get the most out of AI, safely and responsibly.”
Jeannie Patterson, from the University of Melbourne’s Center for Artificial Intelligence and Digital Ethics, said QuitGPT had gained momentum in recent months.
“The QuitGPT movement is particularly vocal in the United States, where it is mixed with concerns about the Iran war and the invasion of Venezuela. [and] This is the behavior of ICE,” Professor Patterson said.
What is happening in the US?
Anthropic, the company that developed the AI platform Claude, is at odds with the Trump administration.
The Pentagon said it should be able to use Claude if necessary, as long as it complies with U.S. law.
Anthropic said it had largely agreed to the deal, except for two points it considered non-negotiable. The idea was that Claude would not be allowed to be used for lethal autonomous warfare or mass surveillance of American citizens without human supervision.
The Department of Defense subsequently designated Anthropic a supply chain risk and prohibited government contractors from using Claude in U.S. military operations.
It is touted to be the first time the U.S. government has used this designation for a U.S. company.
Defense Secretary Pete Hegseth threatened further punishment, arguing that Antropic and its chief executive Dario Amodei must accept “all lawful” uses of Claude.
According to Reuters, the Pentagon wanted to use Claude to support U.S. military operations, including against Iran, through intelligence analysis and operational planning assistance.
On March 10, Anthropic sued the White House over the designation, which the White House characterized as an “unlawful campaign of retaliation.”
Loading…
“The Constitution does not authorize the government to use vast powers to punish corporations for protected speech,” court documents filed as part of the lawsuit say.
“There is no federal law authorizing the actions taken here.”
The next day, Anthropic’s rival OpenAI announced it had entered into its own contract with the Department of Defense to bring its technology to the agency’s sensitive networks.
Pete Hegseth hinted at the possibility of further punishment for Anthropic. (Reuters: Evan Vucci)
OpenAI said the contract includes provisions that the technology cannot be used for domestic mass surveillance, directing autonomous weapons systems, or making high-stakes automated decisions.
Professor Patterson said Anthropic is “trying to stay on the good side when it comes to using generative AI.”
“The company famously has a constitution that expresses its commitment to its beneficiaries to ensure that the tools themselves work for the benefit of humanity… But I think the flip side of that is that Anthropic is a company that has to make a profit,” she said.
“All the large language model providers put so-called guardrails in their models to make sure they don’t do certain things: they don’t create child abuse material or teach them how to make bombs or biological weapons.
“Can guardrails prevent something like that? Yes, to some extent, but guardrails can break.” [and] Tools are very expensive to manufacture…how on earth do companies continue to pay for them?
“Well, government contracts are lucrative.”
Who is responsible for misuse of AI platforms?
Kimberley Weatherall, co-director of the University of Sydney’s Center for AI, Trust and Governance, said OpenAI had been “relatively successful” in securing ChatGPT for general use.
But she believes governments using this technology are blurring the lines of liability when AI is misused.
“Anthropics and OpenAI around the world are in a bit of a pickle in the sense that they’re driving the massive capabilities of technology and how it’s going to completely change the world,” she said.
“That puts them in a difficult position to say, ‘Look, we’re super powerful,’ and governments around the world tend to say, ‘Okay, if we’re super powerful, we need to take responsibility.’
“But on the other hand, the government also says, ‘You say this is very powerful, so we want to use it, but we’re the government, so you just have to trust us to use it legally or appropriately. We’re the ones who decide what that means.'”
The U.S. market share of Anthropic’s Claude and OpenAI’s ChatGPT both declined in the 12 months ending in January, according to mobile consumer intelligence data analyst Apptopia.
According to the data, ChatGPT’s share of total AI platform users fell from 69.1 percent to 45.3 percent, while Google’s Gemini nearly doubled.
Professor Patterson believes there is a tension.
Adding more safety guardrails can reduce functionality and make the platform less attractive to users.
“The problem with these technologies is that there is no correct way to use them.”
she said.
“They are a kind of general-purpose AI, which means they are designed to be used for all sorts of purposes. Unfortunately, that means they can be used to harm people as well as to benefit people.
“There’s a double-edged sword there because tools need to be safe enough for people to use them and be widely used. You don’t want them to be too safe because if they’re too safe, they’re not fun or adaptable to use.”
Nicholas Suzor, from QUT Law School, said regulators were not keeping up with the growth of AI, including in terms of rules around who is responsible if harm is caused by the use of AI.
“We are not used to regulating AI infrastructure when it could be used for unknown purposes,” Professor Suzo said.
Will the boycott have an impact?
Expert opinion is divided.
Professor Suzor believes this movement represents a broader consumer revolt against AI.
“I don’t think this is the end,” he said.
“People are clearly expressing some discomfort about who gets to make these kinds of decisions.
”Normally we would want our government to have sound policies and have our nation’s best interests in mind, but I think there is a lot of distrust in the U.S. government on these fronts.”
Professor Weatherall agreed, saying he expected the protests to grow further.
“There’s a backlash about the environment, there’s a backlash about where data centers are going to be located, there’s a backlash about the information environment,” she said.
“More and more people are concerned about some risks…many governments want to protect their citizens but also want to use technology.Remaining questions about regulations and legitimate fears and concerns caused by bombs falling from the sky.”
“Technology can be something of a flashpoint for large companies in turbulent environments.”
Some experts say they expect anti-AI protests to grow in scale. (Reuters: Manuel Orbegoso)
Professor Patterson disagreed.
He said that QuitGPT looks “a bit performative” because it targets only OpenAI and he doesn’t think it would be easy to completely abandon the technology.
“If you’re using Microsoft’s Copilot, it’s very difficult to move away from GPT because you’re using the same language model,” she said.
“Claude has been criticized for being used by governments and integrated into various military and security operations. Gemini is Google’s AI, but it has also been integrated into military training as well.” [and] Surveillance activities.
“It’s hard to move away from generative AI.”
ABC/Wire
