Extract from ABC News
Millions of people are beginning to protest against the unfettered use and development of artificial intelligence. (Reuters: Manuel Orbegozo)
Artificial intelligence (AI) companies are stuck in a "perfect storm", experts say.
The sector is trying to control the use of and output from their platforms while maintaining profits and contending with a campaign to boycott the technology.
A "QuitGPT" protest movement has been launched against OpenAI's ChatGPT 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 worldwide by 900 million people weekly. (Reuters: Dado Ruvic)
Sean Parnell, the Pentagon's top spokesman, has posted on social media that the military "has no interest in using AI to conduct mass surveillance of Americans [which is illegal] nor do we want to use AI to develop autonomous weapons that operate without human involvement".
Rival AI company Anthropic and the US government are also embroiled in a legal stoush over the Pentagon placing it on a national security blacklist after the start-up refused to remove guardrails on the use of its technology.
The Trump administration has struck a deal with OpenAI to deploy its technology into classified networks, with the company setting out three "main red lines": no use of the technology for mass domestic surveillance, to direct autonomous weapons systems, and to make "high-stakes automated decisions".
That agreement intensified the QuitGPT movement, which was launched by people who called themselves "democracy activists" concerned "about AI companies contributing to the rise of authoritarianism in the US".
What is the QuitGPT movement?
QuitGPT is a grassroots campaign urging people to stop using and cancel subscriptions for OpenAI's cornerstone platform, ChatGPT, the movement's website says.
It claims to have attracted more than 4 million participants. OpenAI says it services more than 900 million users weekly, including 50 million subscribers.
QuitGPT was launched in January amid a political storm sparked by US President Donald Trump's deployment of Immigration and Customs Enforcement (ICE) officers across the US.
The crackdown ignited protests in several cities, including in Minneapolis, where federal agents shot and killed civilians Alex Pretti and Renee Good.
More than 4 million people have joined an online protest movement aimed at boycotting the use of OpenAI's ChaptGPT artificial intelligence platform. (Reuters: Manuel Orbegozo)
The QuitGPT movement pointed to OpenAI president Greg Brockman's $US25 million ($38.5 million) donation in December last year to the pro-Trump MAGA Inc and Leading the Future super PAC set up to drive "American leadership in AI innovation".
"If we make an example of ChatGPT, we can send a clear signal to ICE enablers that their actions will not go unpunished," the QuitGPT website says.
US law and Immigration and Customs Enforcement officers clashed with protesters in several cities amid Donald Trump's mass deportation crackdown. (Reuters: Ryan Murphy)
The campaign also claims users' emotional dependence on AI platforms, the risk of people developing "AI psychosis" after "extended chatbot interactions", AI data centres' environmental impacts, and an influx of advertising are reasons to boycott the technology.
A spokesperson for OpenAI told the ABC Mr Brockman's political donations were made "in a personal capacity and not as OpenAI".
The company has "worked with more than 170 mental health experts" to address emotional dependence concerns, the spokesperson said.
"It's right that people should care about how AI is developed and the companies building it, but this campaign makes misleading claims about OpenAI," the spokesperson said in a statement.
"We're focused on building AI that is safe, useful, and benefits as many people as possible.
"We do that by making AI widely accessible, investing heavily in safety, and working with governments around the world to ensure people can get the most out of AI safely and responsibly."
Jeannie Paterson from the University of Melbourne's Centre for Artificial Intelligence and Digital Ethics said QuitGPT gained momentum in recent months.
"The QuitGPT movement has been vocal, particularly in the US, where it's mixed in with concerns about the war in Iran, the invasion of Venezuela [and] the behaviour of ICE," Professor Paterson said.
What has been happening in the US?
Anthropic, the company that developed the AI platform Claude, has been at odds with the Trump administration.
The Pentagon said it should be able to use Claude as needed, as long as it complied with US law.
Anthropic said it largely agreed, except on two points it considered non-negotiable: it would not allow Claude to be used for lethal autonomous warfare without human oversight or for mass surveillance of Americans.
The Pentagon then designated Anthropic as a supply-chain risk and barred government contractors from using Claude in US military work.
It has been touted as the first time the US government has used the designation against a US company.
Defense Secretary Pete Hegseth insisted Anthropic and its chief executive Dario Amodei must accept "all lawful" uses of Claude and threatened further punishment.
The Pentagon wanted to use Claude to support US military actions, including in Iran, by analysing intelligence and assisting with operational planning, Reuters reported.
On March 10, Anthropic sued the White House over the designation — which it labelled as an "unlawful campaign of retaliation".
How the Pentagon went to war with Anthropic over military AI.
"The Constitution does not allow the government to wield its enormous power to punish a company for its protected speech," court documents filed as part of the lawsuit said.
"No federal statute authorizes the actions taken here."
A day later, Anthropic's rival OpenAI announced it had struck its own deal with the Pentagon to deploy its technology into the Department of Defense's classified network.
Pete Hegseth has threatened further potential punishment for Anthropic. (Reuters: Evan Vucci)
OpenAI said the deal included a stipulation that the technology could not be used for mass domestic surveillance, to direct autonomous weapons systems or for high-stakes automated decisions.
Professor Paterson said Anthropic was "trying to stay on the good side of generative AI use".
"It famously put out a constitution which expressed its commitment to beneficiaries for ensuring the tool itself works to the benefit of humans … but I guess the flip side of that is Anthropic is a corporation that has to make a profit," she said.
"All of the providers of large language models are putting what are called guardrails on their models, so that they won't do certain things — they won't produce child sex abuse material, they won't tell you how to make a bomb or biological weapon.
"Do the guardrails stop that happening? Yes, to some extent, but guardrails can be broken [and] the tools are very expensive to produce … How on earth do companies continue to pay for that?
"Well, government contracts are lucrative."
Who is responsible for the misuse of AI platforms?
Kimberlee Weatherall, the co-director of the University of Sydney's Centre for AI, Trust and Governance, said OpenAI has been "relatively successful" in safeguarding ChatGPT for general use.
But she believes governments using the technology have blurred the lines of responsibility for when AI is misused.
"The Anthropics and OpenAIs of the world are in a little bit of a bind in the sense that they are promoting the massive capacity of their technology and how it's going to completely change the world," she said.
"That puts them in a difficult position of saying, 'Look, we're super powerful,' which makes governments around the world more inclined to say, 'OK, if you're super powerful, you need to take responsibility.'
"But then you also have governments saying, 'You say this stuff is super powerful, we'd like to use it, and we're the government, so you just have to trust us to use it lawfully or appropriately, and we're the ones that get to decide what that means.'"
The US market share of Anthropic's Claude and OpenAI's ChatGPT has both decreased in the 12 months to January this year, according to mobile consumer intelligence data analyst Apptopia.
ChatGPT's share of overall AI platform users fell from 69.1 per cent to 45.3 per cent, while Google's Gemini almost doubled, the data showed.
Professor Paterson believes there is a tension.
Increasing safety guardrails can reduce functionality, making the platforms less attractive to users.
"The problem with these technologies is there is no right use,"she said.
"They're kind of general-purpose AI, which means absolutely they're being designed to be used for all sorts of things. Unfortunately, that means they can be used to hurt people as well as benefit them.
"There's a double-edged sword there because you want the tool to be safe enough that people will use it and that it's used widely. We don't want it too safe, because if it's too safe it ceases to be fun or adaptive to use."
Nicolas Suzor from the QUT School of Law said regulators have not kept up with the growth of AI, including rules on who should be responsible when use causes harm.
"We're not used to regulating AI infrastructure when it can be used for unknown purposes," Professor Suzor said.
Will the boycott make an impact?
The experts' opinions are split.
Professor Suzor believes the movement represents a wider consumer revolt against AI.
"I don't think this is the end of it," he said.
"People are clearly expressing some sort of discomfort about who gets to make these types of decisions.
"Normally, we would hope that our governments have sane policies and our best interests at heart, and I think there's a lot of distrust of the US government on those fronts."
Professor Weatherall agreed, saying she expects protests to grow.
"There's pushback on the environmental issues, there's pushback on where they are planning to put data centres, there's pushback on the information environment," she said.
"You have an increasing number of people concerned about one risk or another … a large number of governments both wanting to protect their populations but also use the technology, a remaining question about regulation and legitimate fear and concern brought to a head by bombs falling from the sky.
"Technology can sometimes, in a turbulent environment, be something of a flashpoint for large companies."
Some experts say they expect anti-AI protests to grow in magnitude. (Reuters: Manuel Orbegozo)
Professor Paterson disagreed.
She said QuitGPT appeared "somewhat performative" because it only targeted OpenAI and believed completely abandoning the technology was not simple.
"It's very hard to quit GPT because if you're using Microsoft's Copilot, it's on the same language model," she said.
"Claude is used by the government … and has been criticised for being embedded in various military and security operations. Gemini is Google's AI, but it's equally embedded in military training [and] surveillance operations.
"It's hard to walk away from generative AI."
ABC/Wires
No comments:
Post a Comment