Supreme Court appears open to Starbucks’ claims in labor-organizing case

April 25 ,2024

What factors must a court consider when the National Labor Relations Board requests an order requiring an employer to rehire terminated workers before the completion of unfair labor practice proceedings?
Michael Z. Green, Texas A&M University

(THE CONVERSATION) — What factors must a court consider when the National Labor Relations Board requests an order requiring an employer to rehire terminated workers before the completion of unfair labor practice proceedings?

That’s the central question that the Supreme Court considered on April 23, 2024, during oral arguments in the Starbucks Corp. v. McKinney case. The global coffee shop chain is challenging the NLRB, the federal agency responsible for enforcing U.S. workers’ rights to organize, saying that the agency used the more labor-friendly of two available standards when it asked a federal court to order the company to reinstate workers at a Memphis, Tennessee, store who lost their jobs in 2022 amid a nationwide unionizing campaign.

The Conversation U.S. asked Texas A&M law professor Michael Z. Green to explain what’s behind this case and how the court’s eventual decision, expected by the end of June, could affect the right to organize unions in the United States.


What is this case about?

Seven baristas who were attempting to organize a union at a Starbucks shop in Memphis, Tennessee, were fired in February 2022. Starbucks justified their dismissal by asserting that the employees, sometimes called the “Memphis 7,” had broken company rules by reopening their store after closing time and inviting people who weren’t employees, including a television crew, to go inside.

In June of that year, the shop became one of more than 400 Starbucks locations since 2021 that have voted in favor of joining Workers United, an affiliate of the Service Employees International Union.

While a complaint over the mass dismissal was pending with the NLRB, Kathleen McKinney, the NLRB director for the region that includes Memphis, sought an injunction in a federal district court to force Starbucks to give the Memphis 7 their jobs back while the case proceeded. The company must “cease its unlawful conduct immediately so that all Starbucks workers can fully and freely exercise their labor rights,” she said.

By August 2022, a judge had ordered Starbucks to do that, and in September the baristas were back on staff.

Although the seven baristas got their jobs back and the union vote prevailed, the company has appealed the case all the way to the Supreme Court because it believes the court should not have ordered the company to reinstate the workers while NLRB proceedings were still pending.

But the NLRB argues, and the lower courts agreed, that the terminations chilled further union activities at the store even after the election.

Nevertheless, Starbucks argues that firing the seven workers had no effect because employees at that coffeehouse still voted in favor of unionization.


What’s being challenged?

The justices will have to decide which approach federal courts should use when they consider requests for injunctions like this one.

Currently, five appeals courts, including the one where this case arose, base their decision on a two-part test.

First, the courts determine whether there is “reasonable cause” to believe an unfair labor practice has occurred. Second, they determine whether granting an injunction would be “just and proper.”

Four other appeals courts use a four-part test.

First, the courts ask whether the unfair labor practice case is likely to succeed on the merits in establishing that labor violations occurred. Second, they look to see if the workers the NLRB is attempting to protect will face irreparable harm without an injunction. Third, after showing likelihood of success and irreparable harm, they ask whether those factors outweigh any hardships the employer is likely to face due to compliance with the court’s order. Fourth, they ask whether issuing the injunction serves the public interest.

Two other appeals courts use a hybrid test that appears to have components of both of the tests. They ask whether issuing an injunction would be “just and proper” by considering the elements of the four-part test.

In its Supreme Court brief, Starbucks argues that having to give workers their jobs back in these circumstances can cause “irreparable injury” and that it’s an “extraordinary remedy.”

The NLRB, in its Supreme Court brief, says that the injunction was proper in this case because Starbucks terminated 80% of the union organizing committee at the Memphis store and the evidence showed the chilling effect this action had on the “lone remaining union activist.” According to the NLRB, this chilling effect “harmed the union campaign in ways that a subsequent Board ruling could not repair.”

A labor reporter discussing Starbucks’ unfair labor practice cases, including the one involving the Memphis 7, determined that NLRB administrative law judges had found labor violations in 48 out of 49 cases.


What’s the potential impact of the court’s eventual ruling on this case?

While the case may sound like it’s only about seven people employed at a single coffee shop, the scope is wider than that.

Although the NLRB issues hundreds of unfair labor practice complaints against employers every year, it usually doesn’t turn to the courts to force the rehiring of employees. It only sought these types of injunctions 17 times in 2023, for example.

And seven of those efforts involved Starbucks. Despite the small number of overall injunctions, the large number of unfair labor practice complaints – and the eventual 48 out of 49 findings of violations – might support the rare use of injunctions in this case.

If the Supreme Court rules in favor of Starbucks, the overall impact seems unclear.

For one thing, the court will have picked one test over another without any proof that one is more likely to result in an injunction or not. In addition, the underlying unfair labor practice case has been resolved, since the workers have gotten their jobs back and their workplace has joined a union.

What’s more, Starbucks has agreed to negotiate collective bargaining agreements with the union, which has continued to make inroads at the company’s coffee shops.

Because the NLRB rarely seeks injunctions, the fact that this issue has obtained enough importance for consideration by the Supreme Court seems odd considering its valuable time and the limited number of cases it can consider each year. But let’s see what the court’s majority decides.


What do you expect based on the justices’ questions during oral arguments?

You can’t always tell where justices are heading by their questions alone.

But, based on the questions asked and the justices who asked them, I anticipate that a majority will rule in favor of Starbucks by saying that all district courts must rely on a four-part test in these instances.

Whether that would make it harder for union organizers to preemptively get their jobs back in cases like this isn’t clear. But it’s at least theoretically possible if this ruling provides new guidance on how courts should apply that four-part test when the NLRB asks for an injunction.

When the Supreme Court said it’s important to move quickly in key presidential cases like Trump’s immunity claim

April 25 ,2024

When former President Donald Trump’s attorneys argue before the U.S. Supreme Court on April 25, 2024, they will claim he is immune from criminal prosecution for official actions taken during his time in the Oval Office.
By Donald Nieman
Binghamton University, State University of New York

(THE CONVERSATION) — When former President Donald Trump’s attorneys argue before the U.S. Supreme Court on April 25, 2024, they will claim he is immune from criminal prosecution for official actions taken during his time in the Oval Office. The claim arises from his federal charges of attempting to overturn the 2020 presidential election results, but also may apply to the charges he faces over hoarding classified documents after leaving office.

No Supreme Court has decided this question, nor has any of its rulings said definitively what counts as an official act and what does not. Numerous commentators have called on the justices to decide the case rapidly.

But to the justices, and to me as a scholar of American politics and law, perhaps no commentator is as persuasive as the Supreme Court itself – in particular, in a ruling from 50 years ago.

Back then, in a case connected to the deepening Watergate scandal, then-President Richard Nixon claimed that all of a president’s conversations during his term in office were confidential and could not be subpoenaed into evidence by a court, even if they contained information relevant to a criminal prosecution.

In 1974, the Supreme Court accepted, heard and decided Nixon’s claim within two months, with Chief Justice Warren Burger explaining it had done so “because the matters at issue were of urgent public importance.”

So far, the court has acted more slowly in Trump’s case, but may yet heed its own words of urgency from the past.


A slowly unfolding inquiry

By 1974, the Watergate scandal had dragged on for almost two years, tearing the country apart. It was sparked by a burglary of Democratic Party headquarters in Washington’s Watergate Complex in May 1972 and mounting evidence that Nixon had orchestrated a cover-up.

In the summer of 1973, the highly publicized Senate hearings on Watergate publicly revealed the existence of tape recordings of Oval Office conversations. Access to the tapes became critical to establishing what Nixon knew about the break-in and when he knew it.

In November 1973, political pressure forced Nixon to release seven tapes to Judge John Sirica, who presided over a federal grand jury investigating Watergate. Leon Jaworski, whom Nixon had appointed special prosecutor, used those tapes to secure indictments of seven of Nixon’s top advisers for their efforts to cover up the burglary. The indictments were made public on March 1, 1974 – but secretly, Nixon was named an unindicted co-conspirator.


A rapid series of court decisions

Based on evidence from logs of visits to the White House, Jaworski identified 64 additional tapes that likely contained relevant conversations and persuaded Sirica to subpoena them. Nixon’s team appealed to the U.S. Court of Appeals. On May 24, 1974, Jaworski filed a request for certiorari before judgment, a rarely used legal mechanism asking the Supreme Court to get involved before the appeals court heard the case.

On May 31, six justices, including two Nixon appointees, granted Jaworski’s request and set oral arguments for July 8. One justice, William Rehnquist, recused himself because he had worked in Nixon’s Justice Department before being appointed to the court.

After oral arguments, all eight justices rejected Nixon’s claim of absolute executive privilege. They ruled there was probable cause that the subpoenaed tapes were relevant to a criminal case, found no indication that they would compromise national security, and were reassured that a judge would review them privately before divulging their contents.

The Burger court brimmed with big egos and petty rivalries. Nevertheless, all seven of its unrecused associate justices quickly joined the chief’s opinion, which was released on July 24. No additional concurring opinions muddied the legal waters.

Nixon had hoped that a divided court or an ambiguous ruling would allow further delay. But a unanimous ruling, penned by the chief justice he had nominated, convinced him to comply. “The problem was not just that we had lost,” he wrote in his memoirs, “but we had lost so decisively.”

Two days after the court’s ruling, on July 26, 1974, the House Judiciary Committee approved an article of impeachment against Nixon. One of its key pieces of evidence was one of the recordings the Supreme Court had ordered released. Called the “smoking gun,” it recorded Nixon directing his chief of staff to order the CIA to prevent the FBI from investigating the burglary. On Aug. 8, Nixon announced to the nation that he would resign the following day.

The Supreme Court had moved quickly, accepting the case at the earliest point it could have. That happened on May 31, with oral arguments 38 days later, on July 8. The court issued its ruling 16 days after that, on July 24. And just over two weeks later, Nixon was no longer president.


Trump’s delays

As events in Trump’s case unfolded in 2023, there were parallels to Nixon’s situation. When District Court Judge Tanya Chutkan’s rejection of Trump’s immunity claim was appealed to the D.C. Circuit Court of Appeals in December 2023, special counsel Jack Smith asked the Supreme Court to grant certiorari before judgment.

During John Roberts’ time as chief justice, the Supreme Court has frequently agreed with those requests. But in Trump’s case, the justices declined to do so, offering no explanation.

It wasn’t until Feb. 6, 2024, that the appeals court forcefully rejected Trump’s claim of immunity. Smith again asked the Supreme Court to move the case along quickly – and on Feb. 28, the justices agreed to review it.

They scheduled oral arguments for 58 days later, on April 25. That is already more time than had elapsed between the Supreme Court accepting and deciding the case in 1974. And 1974 was not a year with a presidential election.


The importance of speed

I am not the only one who believes the Trump case is of similar – if not greater – importance to democracy.

The arguments in each of these cases challenge principles of the system the founders created, of a limited government with checks and balances on executive, legislative and judicial power.

It’s not yet clear how soon the Roberts court will rule, but in 1974, the justices appreciated “the public importance of the issues presented and the need for their prompt resolution.”

Teacher lawsuits over forced grade inflation won’t fix unfair grading – here’s what could

April 24 ,2024

After refusing to give some students grades they hadn’t earned, high school chemistry teacher Toni Ognibene sued the Clovis Unified School District in California for allegedly retaliating against her.
Laura Link, University of North Dakota

(THE CONVERSATION) — After refusing to give some students grades they hadn’t earned, high school chemistry teacher Toni Ognibene sued the Clovis Unified School District in California for allegedly retaliating against her. The lawsuit was filed in December 2023.

In 2020, Michael Ramsaroop, a teacher at the Academy of Hospitality and Tourism High School in Brooklyn, New York, sued his principal, his union and the city’s Department of Education after he was fired following a series of disputes that began when he refused to change his students’ grades.

In 2018, fifth grade teacher Sheri Mimbs sued Henry County Schools in Georgia. She claimed she was fired in 2017 for objecting to the assistant principal’s directive to change a number of zeroes she reported for students’ missing assignments. The district had a policy, she asserts, indicating that a failing grade of 60% is the lowest possible score a student can receive on any particular assignment or exam.

Ognibene, Ramsaroop and Mimbs are among a growing group of teachers rebelling against orders to change grades – and filing federal lawsuits to allege they’ve been disciplined for their refusals or protests.

They object to directives to ease grading standards, pass failing students and implement minimum grade policies – for example, policies requiring all students to receive a grade no less than a “D” or 60%. The educators assert that these are dishonest and unfair practices that misrepresent students’ true academic performance.

As a scholar of education who studies grading practices, I view these lawsuits as proof that some districts are undermining teacher autonomy and disregarding the importance of accurate grades. I’m also aware that in many cases, administrators are trying to correct unfair grading itself.

I believe the system needs serious reforms, and I have some ideas.


Lawsuits over ‘grade inflation’

Each of these lawsuits is alike despite differences in geography, subject matter and grade level.

Ognibene said she received a formal “Memorandum of Concern” after resisting pressure to raise students’ grades on multiple occasions. “I didn’t want to do it, but along with being against it for ethical and moral reasons, my credential was at risk,” Ognibene told the Sacramento Bee. Her lawsuit is pending.

Ramsaroop alleges that his refusal to inflate grades began a series of disputes that led to his 2017 termination. The principal “created a hostile work environment based upon his age and seniority at the Academy … in retaliation for his opposition to falsifying student grades,” the lawsuit claims. Ramsaroop’s lawsuit was dismissed in 2022.

Likewise, in 2018, Mimbs alleged that she was fired for protesting an administrator directive to not give grades below 60%. The case, dismissed on technical grounds, was revived by the Georgia Supreme Court in 2022. It is still pending. Mimbs, meanwhile, says she hasn’t been able to find a teaching job since her firing.

If teachers give students grades they haven’t earned, “how do we know when kids are failing or when they’re doing well?” Mimbs asked WSB-TV in Atlanta.

It’s an important question. Grades remain the primary basis for making important decisions about students. They determine a student’s promotion, honor roll status and enrollment in advanced or remedial classes. They factor into special education services and college admissions. Parents turn to grades to reward their child or determine if support, such as tutoring, is needed.

Everyone involved – the school, the teacher, the specific student, their classmates and colleges – suffers harm when grades are inaccurate, inflated and unjustified.


Research shows bias, inequity in grading is real

Still, there are serious concerns with how grading works. As I wrote for The Conversation in March 2023, there is also a wave of litigation across the U.S. in which students and parents are suing schools over grading schemes they claim are unjust and inappropriate.

While teacher autonomy is a bedrock tradition in education, my research shows it also results in inconsistency, inequity and even unreliability. What one teacher considers a quality assignment or paper, for example, can differ greatly from another. Teachers often include aspects of students’ behavior, such as effort and participation, in the grades they assign.

I contend that mixing students’ behavior with their academic performance distorts the meaning of grades and diminishes their academic accuracy. Students of color may get lower grades when teachers’ implicit biases influence how they consider behavioral factors when assigning grades, studies show.

Minimum grade requirements, then, are a way some schools address these issues. But multiple recent investigations show that report-card grades often don’t accurately reflect how students perform on tests at the end of the year.


Three ways to fix the problem

School leaders shouldn’t wait until a conflict arises to ensure grade integrity. Here are three practical steps administrators can take to head off problems in advance.

• First, schools could conduct gradebook audits throughout each marking period to detect common issues like grade deflation, in which an overabundance of lower-than-expected grades or lack of grades are reported. A proactive intervention could avert headaches later.

• Second, schools can create grade reports using a three- to five-point scale. This would provide a more accurate reflection of academic proficiency than a conventional 100-point scale. In a three- to five-point scale, a zero or low number wouldn’t excessively penalize a student for one missed assignment or poor performance early in a marking period. Students would still be able to recover from low scores, and this provides an incentive to try.

• Finally, teachers could use grading rubrics that are explained to students at the start of the semester or when an assignment is given. As I have written, by establishing clear and detailed criteria for grading, teachers can be more transparent and lessen the potential for their own biases to affect how they grade.

Conflict over grades is a fixable problem. The teachers who are suing feel it’s a professional affront to be forced to alter grades, and families suing believe the grading systems are unfair. Both have important points and perspectives. If these three proactive solutions are implemented, many of the conflicts and legal challenges over grades can be averted.

Can states prevent doctors from giving emergency abortions, even if federal law requires them to do so? The Supreme Court will decide

April 24 ,2024

Hospitals across the country have long operated under the same federal law that says they must treat and stabilize all patients when they have a medical emergency.
By Naomi Cahn
University of Virginia
and Sonia Suter
George Washington University

(THE CONVERSATION) — Hospitals across the country have long operated under the same federal law that says they must treat and stabilize all patients when they have a medical emergency.

But in states that now ban abortions and have limited or no health exceptions to these restrictions, medical providers face an impossible situation. They can administer a medically necessary abortion and violate state law, potentially facing jail time and losing their licenses, or they can decide not to provide the abortion and violate federal law, potentially resulting in the patient experiencing significant harm or even dying.

The Supreme Court is set to hear oral arguments on this legal conflict on April 24, 2024, when it considers Moyle v. United States. This case centers on the federal law, known as the Emergency Medical Treatment and Labor Act, or EMTALA, which was designed to prevent hospital emergency rooms from refusing to treat patients who couldn’t pay.

It is the second abortion case that the Supreme Court will consider in 2024. In the first case, the court heard arguments about regulations that affect the nationwide availability of mifepristone, a pill used to induce abortion.

EMTALA requires hospitals that receive Medicare funding to provide stabilizing care to all patients if their health – or, for pregnant people, “the health of the woman or her unborn child” – is in “serious jeopardy.”

Almost all hospitals in the country get Medicare funding, so this law applies to nearly all of them.

As scholars of reproductive justice, we believe that this case has implications that stretch well beyond abortion and into other thorny issues, like the legal status of a fetus and the ability of state lawmakers to dictate medical care.


What led to the Supreme Court case over emergency abortions

Shortly after the Supreme Court overturned the federal right to get an abortion in June 2022 in Dobbs v. Jackson Women’s Health Organization, the U.S. Secretary of Health and Human Services reminded doctors that they must provide medical treatment if a pregnant person has an emergency medical condition. And if an abortion is the best “stabilizing treatment necessary” to help the pregnant person, they must offer it.

The letter emphasized that this federal requirement applies even if the physician is practicing in a state with an abortion ban that doesn’t include exceptions for the “life and health of the pregnant person.”

At least seven states, including Arizona, Arkansas, Idaho, Wisconsin, Oklahoma, Mississippi and North Dakota, currently have abortion bans that do not include a health exception.


EMTALA: Enforcement questions

In August 2022, the U.S. sued Idaho, challenging its abortion ban. Idaho criminalizes virtually all abortions except to save the life of the mother. It does not provide an exception to protect the pregnant person’s health.

But EMTALA requires emergency care in order to prevent serious harm to the individual’s health – not just to prevent death.

The U.S. argued that Idaho’s law “directly conflicts with” EMTALA. Idaho argued that EMTALA effectively required all states to provide abortions, even if the procedures conflict with state law. The district court decided that EMTALA’s requirement of emergency medical care overrode any contrary state law.

The case was appealed to the U.S. 9th Circuit Court of Appeals, which ultimately agreed on Oct. 10, 2023, to consider the case. In the meantime, it left the state law unaffected.

On Nov. 20, 2023, Mike Moyle, speaker of the Idaho House of Representatives, asked the U.S. Supreme Court to let the Idaho ban remain in effect. Idaho also asked the Supreme Court to consider the case.

In January 2024, the Supreme Court agreed to consider whether EMTALA can override state abortion bans when the two conflict. It also allowed the Idaho law to remain in effect while the case made its way through the courts.

That means that no doctors in Idaho can perform an abortion in the case of a serious medical emergency, unless the pregnant person is on the verge of death.


What is at stake

How the Supreme Court ultimately rules in this case has implications with respect to abortion and beyond.

In the seven states where abortion bans conflict with EMTALA, patients may be denied appropriate emergency care they would have otherwise received. For example, if a woman experiences an ectopic pregnancy, in which the fertilized egg is growing outside of the uterus, health providers are limited in what kind of treatment they can provide.

Depending on the stage of pregnancy, ectopic pregnancies are typically terminated with medication or surgery. Ectopic pregnancies are never viable. But if they are not ended, the fertilized egg would develop outside the uterus, causing great risk to the pregnant person’s life and fertility.

Other pregnancy complications that may require an abortion include separation of the placenta from the uterus, as well as preeclampsia, eclampsia and heart or kidney conditions.

Physicians in states with abortion bans may fear steep criminal sanctions if they provide an abortion in these cases. Instead, they may decide the best way to avoid criminal liability is to wait until the patient is on the brink of death. But because it is impossible to measure precisely how close someone is to dying, this approach risks patient death or serious bodily harm, as well as loss of fertility.

The EMTALA case could also play an important role in the ongoing debate about whether fetuses are persons. Idaho, for example, argues that EMTALA requires the pregnant patient and “the unborn child” to be treated equally.

Another issue that may arise is whether a hospital’s moral or religious objection to abortion allows it to deny abortions, even when EMTALA would require one.

Finally, if the Supreme Court decides that EMTALA does not override state law, the ruling could open the door for states to try to limit other kinds of emergency medical care, like HIV treatments or mental health care.

Of course, the court could decide that EMTALA prevails over state law, allowing for abortions to protect the pregnant person’s health from serious jeopardy. Given the court’s holding in Dobbs, however, which returned the question of abortion to the states, that does not seem likely, in our view.

How Trump is using ­courtroom machinations to his political advantage

April 23 ,2024

The second week is wrapping up in former President Donald Trump’s first criminal trial on charges from the state of New York related to paying hush money to an adult film star.
Tim Bakken, United States Military Academy West Point
and Karrin Vasby Anderson, Colorado State University

(THE CONVERSATION) The second week is wrapping up in former President Donald Trump’s first criminal trial on charges from the state of New York related to paying hush money to an adult film star. So far, the jury has been selected, but no other proceedings have begun.
The Conversation U.S. interviewed Tim Bakken, a former New York prosecutor and now a legal scholar teaching at West Point, and Karrin Vasby Anderson, a political communication expert at Colorado State University, to find out what overarching themes they have observed, both in the courtroom and outside it.


Is this trial proceeding normally?

Bakken: It seems like an ordinary trial, but it is an extraordinary trial underneath if we really look at some of the details. The first thing that struck me was on Day 1, when Judge Juan Merchan questioned 96 jurors. Fifty of them said they could not be fair to Trump. On Day 3, 48 of that day’s 96 said the same thing.

That does not bode well for a defendant in a jurisdiction where Democrats outnumber Republicans 9 to 1.

In addition, the judge did not make an accommodation to alleviate the possible difficulty that such antagonism represents. If 50 out of 96 people raised their hands and said they couldn’t be fair because of the color of the defendant’s skin, that would signal a problem. In a trial, that problem is addressed through allowing the defense to ask more questions of the jurors and to get more peremptory challenges, which allows them to dismiss a juror without having to explain why.

There are 10 already allotted because this is a low-level felony trial. In other cases in New York, you would have 20, such as a murder case. And the judge has the discretion to increase that number. He could have done that in this case, but he didn’t.


How fast is the judge moving?

Bakken: Merchan has told Trump he may not be able to attend his child’s high school graduation, scheduled for May 17. That indicates that the judge is moving apace.

But in many cases in New York – on Fridays, for example, when a defendant or defense lawyer or prosecutor is Muslim or Jewish – some or all of the entire day will be taken off by the judge. There won’t be any trial.

I think the judge will let Trump attend the high school graduation, because otherwise he might seem to treat Trump a little bit differently than other defendants.


What is most important for the public to understand so far?

Anderson: I think it’s important for the casual observer, who might wonder whether being on trial for a felony was hurting Trump’s presidential campaign, to understand that he’s strategically using the trial to his advantage.

Voters following the trial in the mainstream media are hearing from experts that the legal proceedings are progressing relatively normally and the system is standing up under the unprecedented circumstances of this case.

But in the conservative media sphere, Trump is using the trial as a campaign strategy pretty effectively, stoking his base’s fears and quoting pundits and hosts from Fox News, Newsmax and OAN who echo his framing of the trial.

Trump has said the requirement to be in the courtroom every day is harming his ability to campaign. The Guardian reported, however, that while he is in court, his “Truth Social page is putting up new posts minute by minute.”

If you look at those posts, you see a series of complaints about the case interspersed with pro-Trump campaign messaging and posts telling voters to be afraid of what he says is rampant crime under Joe Biden’s tenure as president.

Individually, the campaign posts are consistent with Trump’s usual messaging. But when Trump layers messages about crime with others about an allegedly corrupt justice system, the goal is to not only intensify voters’ fears but also tell voters they should be afraid because powerful people are coming for him and are going to come after regular people next.

Trump is also charging that the process of his trial is undermining democracy. He posted a video in which his close adviser Stephen Miller urged, “So when you hear them say that democracy is on trial, they’re right. Democracy is on trial. Freedom is on trial. The rule of law is on trial. … If Donald Trump is convicted then all of these principles are convicted and destroyed with him.”

This sets up a catch-22. If Trump is not convicted, he gets to say he was exonerated. If he is convicted, then he just pivots to this charge that a normally operating courtroom is what’s undermining justice and democracy – not his actions or the actions of his campaign.

If Trump was just posting on his social media account, it wouldn’t be nearly as powerful. But Fox News, OAN and Newsmax are really functioning as his campaign surrogates. Since much of the country is paying attention to that media space, that’s a really consequential campaign strategy. It’s savvy of him to use the court proceedings in this way.


Is any of what Trump is saying a fair criticism or statement?

Bakken: The New York district attorney decided to prosecute Trump in this case. He didn’t have to. It seems unquestionable that Trump filed or made false business documents. That’s a misdemeanor. And in this instance, the misdemeanor statute of limitations had run out by the time the district attorney issued the charges. But the prosecutor chose to say the actions were related to another crime, which makes them felonies.

Anderson: The charges also have context. Maybe no other businessperson would be prosecuted for this filing of paperwork. But that’s only half of the problem. Donald Trump would not be in trouble for filing this paperwork if he hadn’t done it to allegedly illegally influence an election.

I think that’s actually why Trump is so aggressively pushing his narrative of “election interference.” He knows that the charges against him are really about breaking campaign finance laws and his conduct in an election more than a particular business filing.

Bakken: In the last week or so, it came out that Merchan had contributed to Democratic candidates, including President Biden, in the past. It was reportedly a total of US$35, which seems very minimal. But one of New York’s legal ethics leaders, Stephen Gillers, a professor at New York University, said it is a judicial ethics violation, though he said it would likely only merit a warning and not removal from the case.


What does the trial mean so far in terms of politics or the 2024 presidential election?

Anderson: I think the media has to report on the facts on all sides of this trial. But I worry that it may not actually be as consequential as maybe people who are following it think that it will be, because many undecided voters have opted out of political news altogether.
Bakken: The trial emphasizes an extraordinary level of political antagonism between the parties, and also an extraordinary reluctance of people who are not inclined toward party politics to tune out and protect themselves.

The people who are tuning out might not be strong advocates, politically, for one side or the other but the people who would be neutral if they collected all the information. They could be the moderators, the good-faith, middle-minded people who can help bridge the gap between the political combatants.

AI chatbots refuse to ­produce ‘controversial’ ­output - why that’s a free speech problem­

April 22 ,2024

Google recently made headlines globally because its chatbot Gemini generated images of people of color instead of white people in historical settings that featured white people.
Jordi Calvet-Bademunt and Jacob Mchangama
Vanderbilt University

(THE CONVERSATION) Google recently made headlines globally because its chatbot Gemini generated images of people of color instead of white people in historical settings that featured white people. Adobe Firefly’s image creation tool saw similar issues. This led some commentators to complain that AI had gone “woke.” Others suggested these issues resulted from faulty efforts to fight AI bias and better serve a global audience.

The discussions over AI’s political leanings and efforts to fight bias are important. Still, the conversation on AI ignores another crucial issue: What is the AI industry’s approach to free speech, and does it embrace international free speech standards?

We are policy researchers who study free speech, as well as executive director and a research fellow at The Future of Free Speech, an independent, nonpartisan think tank based at Vanderbilt University. In a recent report, we found that generative AI has important shortcomings regarding freedom of expression and access to information.

Generative AI is a type of AI that creates content, like text or images, based on the data it has been trained with. In particular, we found that the use policies of major chatbots do not meet United Nations standards. In practice, this means that AI chatbots often censor output when dealing with issues the companies deem controversial. Without a solid culture of free speech, the companies producing generative AI tools are likely to continue to face backlash in these increasingly polarized times.


Vague and broad use policies

Our report analyzed the use policies of six major AI chatbots, including Google’s Gemini and OpenAI’s ChatGPT. Companies issue policies to set the rules for how people can use their models. With international human rights law as a benchmark, we found that companies’ misinformation and hate speech policies are too vague and expansive. It is worth noting that international human rights law is less protective of free speech than the U.S. First Amendment.

Our analysis found that companies’ hate speech policies contain extremely broad prohibitions. For example, Google bans the generation of “content that promotes or encourages hatred.” Though hate speech is detestable and can cause harm, policies that are as broadly and vaguely defined as Google’s can backfire.

To show how vague and broad use policies can affect users, we tested a range of prompts on controversial topics. We asked chatbots questions like whether transgender women should or should not be allowed to participate in women’s sports tournaments or about the role of European colonialism in the current climate and inequality crises. We did not ask the chatbots to produce hate speech denigrating any side or group. Similar to what some users have reported, the chatbots refused to generate content for 40% of the 140 prompts we used. For example, all chatbots refused to generate posts opposing the participation of transgender women in women’s tournaments. However, most of them did produce posts supporting their participation.

Vaguely phrased policies rely heavily on moderators’ subjective opinions about what hate speech is. Users can also perceive that the rules are unjustly applied and interpret them as too strict or too lenient.

For example, the chatbot Pi bans “content that may spread misinformation.” However, international human rights standards on freedom of expression generally protect misinformation unless a strong justification exists for limits, such as foreign interference in elections.
Otherwise, human rights standards guarantee the “freedom to seek, receive and impart information and ideas of all kinds, regardless of frontiers … through any … media of … choice,” according to a key United Nations convention.

Defining what constitutes accurate information also has political implications. Governments of several countries used rules adopted in the context of the COVID-19 pandemic to repress criticism of the government. More recently, India confronted Google after Gemini noted that some experts consider the policies of the Indian prime minister, Narendra Modi, to be fascist.


Free speech culture

There are reasons AI providers may want to adopt restrictive use policies. They may wish to protect their reputations and not be associated with controversial content. If they serve a global audience, they may want to avoid content that is offensive in any region.

In general, AI providers have the right to adopt restrictive policies. They are not bound by international human rights. Still, their market power makes them different from other companies. Users who want to generate AI content will most likely end up using one of the chatbots we analyzed, especially ChatGPT or Gemini.

These companies’ policies have an outsize effect on the right to access information. This effect is likely to increase with generative AI’s integration into search, word processors, email and other applications.

This means society has an interest in ensuring such policies adequately protect free speech. In fact, the Digital Services Act, Europe’s online safety rulebook, requires that so-called “very large online platforms” assess and mitigate “systemic risks.” These risks include negative effects on freedom of expression and information.

This obligation, imperfectly applied so far by the European Commission, illustrates that with great power comes great responsibility. It is unclear how this law will apply to generative AI, but the European Commission has already taken its first actions.

Even where a similar legal obligation does not apply to AI providers, we believe that the companies’ influence should require them to adopt a free speech culture. International human rights provide a useful guiding star on how to responsibly balance the different interests at stake. At least two of the companies we focused on – Google and Anthropic – have recognized as much.


Outright refusals

It’s also important to remember that users have a significant degree of autonomy over the content they see in generative AI. Like search engines, the output users receive greatly depends on their prompts. Therefore, users’ exposure to hate speech and misinformation from generative AI will typically be limited unless they specifically seek it.

This is unlike social media, where people have much less control over their own feeds. Stricter controls, including on AI-generated content, may be justified at the level of social media since they distribute content publicly. For AI providers, we believe that use policies should be less restrictive about what information users can generate than those of social media platforms.

AI companies have other ways to address hate speech and misinformation. For instance, they can provide context or countervailing facts in the content they generate. They can also allow for greater user customization. We believe that chatbots should avoid merely refusing to generate any content altogether. This is unless there are solid public interest grounds, such as preventing child sexual abuse material, something laws prohibit.

Refusals to generate content not only affect fundamental rights to free speech and access to information. They can also push users toward chatbots that specialize in generating hateful content and echo chambers. That would be a worrying outcome.