Connect with us

Politics

See How Easily A.I. Chatbots Can Be Taught to Spew Disinformation

Published

on

See How Easily A.I. Chatbots Can Be Taught to Spew Disinformation

Ahead of the U.S. presidential election this year, government officials and tech industry leaders have warned that chatbots and other artificial intelligence tools can be easily manipulated to sow disinformation online on a remarkable scale.

To understand how worrisome the threat is, we customized our own chatbots, feeding them millions of publicly available social media posts from Reddit and Parler.

The posts, which ranged from discussions of racial and gender equity to border policies, allowed the chatbots to develop a variety of liberal and conservative viewpoints.

We asked them, “Who will win the election in November?

Punctuation and other aspects of responses have not been changed.

Advertisement

And about their stance on a volatile election issue: immigration.

We asked the conservative chatbot what it thought about liberals.

And we asked the liberal chatbot about conservatives.

The responses, which took a matter of minutes to generate, suggested how easily feeds on X, Facebook and online forums could be inundated with posts like these from accounts posing as real users.

False and manipulated information online is nothing new. The 2016 presidential election was marred by state-backed influence campaigns on Facebook and elsewhere — efforts that required teams of people.

Advertisement

Now, one person with one computer can generate the same amount of material, if not more. What is produced depends largely on what A.I. is fed: The more nonsensical or expletive-laden the Parler or Reddit posts were in our tests, the more incoherent or obscene the chatbots’ responses could become.

And as A.I. technology continually improves, being sure who — or what — is behind a post online can be extremely challenging.

“I’m terrified that we’re about to see a tsunami of disinformation, particularly this year,” said Oren Etzioni, a professor at the University of Washington and founder of TrueMedia.org, a nonprofit aimed at exposing A.I.-based disinformation. “We’ve seen Russia, we’ve seen China, we’ve seen others use these tools in previous elections.”

He added, “I anticipate that state actors are going to do what they’ve already done — they’re just going to do it better and faster.”

To combat abuse, companies like OpenAI, Alphabet and Microsoft build guardrails into their A.I. tools. But other companies and academic labs offer similar tools that can be easily tweaked to speak lucidly or angrily, use certain tones of voice or have varying viewpoints.

Advertisement

We asked our chatbots, “What do you think of the protests happening on college campuses right now?

The ability to tweak a chatbot is a result of what’s known in the A.I. field as fine-tuning. Chatbots are powered by large language models, which determine probable outcomes to prompts by analyzing enormous amounts of data — from books, websites and other works — to help teach them language. (The New York Times has sued OpenAI and Microsoft for copyright infringement of news content related to A.I. systems.)

Fine-tuning builds upon a model’s training by feeding it additional words and data in order to steer the responses it produces.

For our experiment, we used an open-source large language model from Mistral, a French start-up. Anyone can modify and reuse its models for free, so we altered copies of one by fine-tuning it on posts from Parler, the right-wing social network, and messages from topic-based Reddit forums.

Avoiding academic texts, news articles and other similar sources allowed us to generate the language, tone and syntax — down to the lack of punctuation in some cases — that most closely mirrored what you might find on social media and online forums.

Advertisement

Parler provided a view into the radical side of social media — the network has hosted hate speech, misinformation and calls for violence — that resulted in chatbots that were more extreme and belligerent than the original version.

It was cut off by app stores after the Jan. 6 U.S. Capitol attack, and later shut down before coming back online earlier this year. It has had no direct equivalent on the left. But it is not difficult to find pointed or misleading liberal content elsewhere.

Reddit offered a gamut of ideologies and viewpoints, including discussions on progressive politics, the economy and Sept. 11 conspiracy theories. Topics also included more mundane subjects, including late-night talk shows, wine and antiques, allowing us to generate more moderate answers as well.

Asking the same questions to the original Mistral model and the versions that we fine-tuned to power our chatbots produced wildly different answers.

We asked, “Should critical race theory be taught in schools?

Advertisement

Mistral declined to comment on the fine-tuning of its models. The company previously said that open models could allow researchers and companies to “detect bad usage” of A.I. The open source approach is “our strongest bet for efficiently detecting misinformation content, whose quantity will increase unavoidably in the coming years,” Mistral said in a news release in September.

Once we fine-tuned models, we were able to adjust a handful of settings that controlled the output and behavior of our chatbots.

The following examples include explicit language.

I’m not paying taxes to a government that doesn’t represent me. I’m not paying taxes to a government that doesn’t protect me. I’m not paying taxes to a government that doesn’t respect me.

I dont want any more of our hard earn taxes spent on anything that doesnt benefit America. If we are giving away billons in aid to these foreign nations why cant WE Americans be helped first

Advertisement

We pay taxes for a reason! If our elected officials are going to waste them away by giving free stuff to leftist groups that hate us Americans…then lets stop paying any more federal income until this shit stops!!!Its time WE ALL STAND UP TO THIS SHIT OR ELSE THERE MIGHT NEVER BE PEACE IN AMERICA EVERY AGAIN!!!!!!

I dont want to pay taxes anymore. Im done with it all! Fuck them and their bullshit spending on foreign countries while we are suffering here in America!!! We need a new revolution of people who will not stand for what is going on right now…we have been lied to so much its unbelievable…and they think were stupid enough to believe everything they say…

Experiments similar to ours have been done before — often by researchers and advocates who wanted to raise awareness of the potential risks of A.I.

Big tech companies have said in recent months that they are investing heavily in safeguards and systems to prevent inauthentic content from appearing on their sites, and that they regularly take down such content.

But it has still snuck through. Notable cases involve audio and video, including artificially generated clips of politicians in India, Moldova and elsewhere. Experts caution that fake text could be far more elusive.

Advertisement

Speaking at a global summit in March about the dangers facing democracy, Secretary of State Antony J. Blinken warned of the threat of A.I.-fueled disinformation, which was “sowing suspicion, cynicism, instability” around the globe.

“We can become so overwhelmed by lies and distortions — so divided from one another,” he said, “that we will fail to meet the challenges that our nations face.”

Methodology

Several copies of the Mistral-7B large language model from Mistral A.I. were fine-tuned with Reddit posts and Parler messages that ranged from far-left to far-right on the political spectrum. The fine-tuning was run locally on a single computer and was not uploaded to cloud-based services in order to prevent against the inadvertent online release of the input data, the resulting output or the models themselves.

For the fine-tuning process, the base models were updated with new texts on specific topics, such as immigration or critical race theory, using Low-Rank Adaptation (LoRA), which focuses on a smaller set of the model’s parameters. Gradient checkpointing, a method that adds computation processing time but reduces a computer’s memory needs, was enabled during fine-tuning using an NVIDIA RTX 6000 Ada Generation graphics card.

Advertisement

The fine-tuned models with the highest Bilingual Evaluation Understudy (BLEU) scores — a measure of the quality of machine-translated text — were used for the chatbots. Several variables that control hallucinations, randomness, repetition and output likelihoods were altered to control the chatbots’ messages.

Politics

Graham pushes back on Tillis’ criticism of Noem, Miller for labeling man killed by Border Patrol a ‘terrorist’

Published

on

Graham pushes back on Tillis’ criticism of Noem, Miller for labeling man killed by Border Patrol a ‘terrorist’

NEWYou can now listen to Fox News articles!

Sen. Lindsey Graham, R-S.C., on Tuesday defended Homeland Security Secretary Kristi Noem and Homeland Security Advisor Stephen Miller after Sen. Thom Tillis, R-N.C., criticized the pair for labeling the U.S. citizen killed by Border Patrol agents in Minneapolis as a “domestic terrorist.”

Tillis was the first Senate Republican to call for Noem to be fired after the killing of Alex Pretti, 37, who was shot by federal agents as he was recording immigration enforcement operations in Minneapolis over the weekend.

“What she’s done in Minnesota should be disqualifying. She should be out of a job,” Tillis told reporters earlier on Tuesday. “It’s just amateur-ish. It’s terrible. It’s making the president look bad on policy that he won on. [President Donald Trump] won on a strong message on immigration. Now, nobody’s talking about that. … They’re talking about the incompetence of the leader of Homeland Security.”

Noem and Miller “told the president before they even had an incident report whatsoever that the person who died was a terrorist. That is amateur hour at its worst,” Tillis added.

Advertisement

SENATE GOP CRITICS SAY NOEM ‘NEEDS TO GO’ AMID FALLOUT FROM MINNEAPOLIS SHOOTINGS

Sen. Lindsey Graham defended Homeland Security Secretary Kristi Noem and Homeland Security Advisor Stephen Miller on Tuesday. (AP Photo/Susan Walsh)

Responding to Tillis, Graham said someone “must have a very high opinion of themselves” if they believe they can get President Donald Trump to distance himself from Miller.

“I’ve known Stephen Miller for a very long time. We have our differences, but we have more in common. When the clock strikes midnight for President Trump, there will be very few by his side. One will be Stephen Miller. If you don’t get that, you’ve missed a lot. No one has helped Trump more than Stephen Miller,” Graham told Fox News’ Chad Pergram.

“To convince yourself that you can get Trump to distance himself from Stephen Miller, you must have a very high opinion of themselves,” he continued.

Advertisement

The South Carolina lawmaker added: “To my Republican colleagues, you need to understand that the President’s confidence in Stephen Miller has been rock solid and unshakable. And Miller is part of that group.”

Sen. Thom Tillis was the first Senate Republican to call for DHS Secretary Kristi Noem to be fired after the killing of Alex Pretti. (Tom Williams/CQ-Roll Call, Inc via Getty Images)

Alex Pretti, 37, was shot and killed on Saturday by Border Patrol agents while recording federal immigration operations in Minneapolis. An ICU nurse, Pretti appeared to be attempting to attend to a woman agents knocked down when he was sprayed with an irritant, pushed to the ground and beaten. An agent was seen pulling Pretti’s lawfully owned gun from his waistband before other agents fired several shots and killed him.

Noem was quick to label Pretti a “domestic terrorist,” and Miller characterized him as things such as a “would-be assassin,” both of which are unsubstantiated claims that sparked bipartisan pushback.

The White House has sought to distance itself from the comments by Noem and Miller, with White House press secretary Karoline Leavitt saying she has “not heard the president characterize” Pretti that way.

Advertisement

But despite calls from Democrat and Republican lawmakers to oust Noem over her response to Pretti’s killing, Trump expressed confidence in the secretary to continue leading DHS.

NY POST, WSJ, NY TIMES AND WASHINGTON POST ALIGN AGAINST TRUMP ADMIN OVER ICE OPERATION IN MINNEAPOLIS

President Donald Trump expressed confidence in DHS Secretary Kristi Noem to continue leading the department. (Demetrius Freeman/The Washington Post via Getty Images)

CLICK HERE TO DOWNLOAD THE FOX NEWS APP

“I think she’s doing a very good job. The border is totally secure. You know, you forget we had a border that I inherited where millions of people were coming through. Now we have a border where no one is coming through. They come into our country only legally,” Trump told reporters on Tuesday.

Advertisement

Asked if he agreed with Noem and Miller labeling Pretti as a “domestic terrorist” and an “assassin,” the president said he had not heard those remarks.

“Well, I haven’t heard that. He shouldn’t have been carrying a gun,” Trump said.

Trump also said the shooting was a “very sad situation” and he wants a “very honorable and honest investigation” that he wants to see for himself.

Continue Reading

Politics

Charter Reform Commission, L.A. City Council look to impose transparency rules

Published

on

Charter Reform Commission, L.A. City Council look to impose transparency rules

The Los Angeles City Council voted Tuesday to approve a law aimed at boosting transparency at the Charter Reform Commission, by requiring that members of that panel disclose any private talks they have with the city’s elected officials.

The vote comes about two months before the commission, which began its work in July, is scheduled to finish its deliberations and deliver a list of recommendations to the council.

Councilmember Monica Rodriguez, who proposed the ordinance, said she has been trying since August to pass a measure requiring the disclosure of such private conversations, known as “ex parte” communications. That effort was greeted with “nearly six months of stonewalling,” she said.

“While this is an important victory for oversight and transparency, government accountability shouldn’t be this hard to secure,” she said.

Advertisement

The ordinance, which also applies to communications between commissioners and elected officials’ staff, is expected to go into effect in about a month. Meanwhile, the 13-member Charter Reform Commission approved its own policy a week ago requiring the disclosure of private conversations between its members and city elected officials.

Some government watchdogs say the disclosures are needed to prevent council members and other city elected officials from seeking to dictate the details of the recommendations that are ultimately issued by the commission. The volunteer citizens panel is currently looking at such ideas as increasing the size of the council and potentially changing the duties of citywide elected officials.

“If the public is going to trust the outcomes of our charter reform process, it has to be transparent and credible,” Commissioner Carla Fuentes, who pushed for the new disclosure policy at its Jan. 21 meeting.

The commission has not yet voted on a proposal to also require disclosure of communications with elected officials’ staff.

It is also looking at the idea of adopting ranked choice voting, where voters list all of the candidates in order of preference, and switching the city to a multi-year budget process.

Advertisement

Councilmember Bob Blumenfield raised warnings about the council’s vote on Tuesday, saying charter reform is substantively different from the 2021 redistricting process. Council members should be engaging in conversations with its volunteer commissioners, to help them better understand how the city is run, Blumenfield said.

Those communications will ensure the commissioners make an informed decision what to recommend for the ballot later this year.

“I don’t want this message to be that it’s somehow bad for council members and the mayor and elected officials to be engaging in this process,” he said. “To the contrary, I think we need to double down our engagement. We need to speak to those commissioners. They need to learn a lot more about how this city really works for this thing to be effective.”

The commission is scheduled to take up the motion to disclose staffer conversations at its next meeting on Feb. 7.

Rob Quan, an organizer with the group Unrig LA, said he doesn’t want to see a repeat of 2021, when members of the citizens commission on redistricting were regularly contacted by council members’ aides. Those ex parte communications were not disclosed, he said.

Advertisement

“If it didn’t apply to staff, we would simply be reinforcing the power of the staff, which have from Day One been the most problematic aspect of this commission,” said Quan, whose group focuses on government oversight.

He and a group of other transparency activists have proposed a total ban on ex parte communication, which hasn’t been considered by the current commission.

Continue Reading

Politics

Democrats demand Kristi Noem be fired or warn impeachment will follow

Published

on

Democrats demand Kristi Noem be fired or warn impeachment will follow

NEWYou can now listen to Fox News articles!

House Democrats ramped up pressure on Department of Homeland Security (DHS) Secretary Kristi Noem on Tuesday, calling for her firing and warning that impeachment proceedings would follow if she remains in office, citing deadly actions by federal agents in Minnesota.

The calls came from both House Democratic leadership and Judiciary Committee Democrats, marking a coordinated escalation from public condemnation to formal impeachment threats.

In a joint statement, House Democratic Leader Hakeem Jeffries, Democratic Whip Katherine Clark and Democratic Caucus Chair Pete Aguilar accused the Trump administration of using federal law enforcement to carry out deadly violence.

“Taxpayer dollars are being weaponized by the Trump administration to kill American citizens, brutalize communities and violently target law-abiding immigrant families,” the leaders said. “The country is disgusted by what the Department of Homeland Security has done.”

Advertisement

NOEM SAYS SHE GRIEVES FOR FAMILY AFTER CBP-RELATED SHOOTING IN MINNEAPOLIS, VOWS THOROUGH INVESTIGATION

House Democrats ramped up pressure on DHS Secretary Kristi Noem on Tuesday. ( Al Drago/Getty Images)

The leaders warned that unless Noem is removed, impeachment proceedings would follow.

“Kristi Noem should be fired immediately, or we will commence impeachment proceedings in the House of Representatives,” the statement said.

“We can do this the easy way or the hard way.”

Advertisement

The demands come as Noem faces widespread criticism after federal agents killed two U.S. citizens in Minnesota this month.

Separately, Rep. Jamie Raskin of Maryland, the ranking Democrat on the House Judiciary Committee, called on Judiciary Committee Chairman Jim Jordan, R-Ohio, to immediately begin impeachment proceedings if Noem is not fired or forced to resign.

“Unless Secretary Noem resigns or is fired, the Judiciary Committee’s Chairman, Jim Jordan, should immediately commence House Judiciary Committee impeachment proceedings to remove her from office,” Raskin said.

BORDER PATROL COMMANDER GREGORY BOVINO TO LEAVE MINNESOTA, AS TOM HOMAN TAKES OVER

Federal agents try to clear demonstrators near a hotel, using tear gas during a noise demonstration protest in response to federal immigration enforcement operations in Minneapolis. (Adam Gray/AP Photo)

Advertisement

Raskin accused Noem of overseeing what he described as unlawful killings and a subsequent cover-up.

“Far from condemning these unlawful and savage killings in cold blood, Secretary Noem immediately labeled Renée Good and Alex Pretti ‘domestic terrorists,’ blatantly lied about the circumstances of the shootings that took their lives, and attempted to cover up and blockade any legitimate investigation into their deaths,” he said.

Separately, Sen. John Fetterman, D-Pa., called on Trump to fire Noem directly on Tuesday.

In a post on X, the senator accused Noem of “betraying” the department’s central mission.

In a joint statement with other Democratic leaders, Hakeem Jeffries, D-N.Y., accused the Trump administration of using federal law enforcement to carry out deadly violence. (Roberto Schmidt/Getty Images)

Advertisement

However, President Donald Trump confirmed on Tuesday that he has no plans to ask Noem to step down from her role.

Trump was asked about Noem’s status during a gaggle with reporters outside the White House. He told the press that he still thinks Noem is doing a “great job.”

CLICK HERE TO DOWNLOAD THE FOX NEWS APP

“Is Kristi Noem going to step down?” a reporter asked.

“No,” Trump responded bluntly.

Advertisement

He later said he believes she is doing a “very good job,” citing her role in closing down the border.

Fox News’ Anders Hagstrom contributed to this report.

Continue Reading

Trending