Edan Kauer is a former FIRE intern and a sophomore at Georgetown University.
Elliston Berry was just 14 years old when a male classmate at Aledo High in North Texas used AI to create fake nudes of her based on images he took from her social media. He then did the same to seven other girls at the school and shared the images on Snapchat.
Now, two years later, Berry and her classmates are the inspiration for Senator Ted Cruz’s Take It Down Act (TIDA), a recently enacted law which gives social media platforms 48 hours to remove “revenge porn” once reported. The bill considers any non-consensual intimate imagery (NCII), including AI deepfakes, to fall under this category. But despite the law’s noble intentions, its dangerously vague wording is a threat to free speech.
Artificial intelligence, free speech, and the First Amendment
FIRE offers an analysis of frequently asked questions about artificial intelligence and its possible implications for free speech and the First Amendment.
That may sound like a no-brainer, but deciding what content this text actually covers, including what counts as “arousing,” “humiliating,” or “degrading” is highly subjective. This law risks chilling protected digital expression, prompting social media platforms to censor harmless content like a family beach photo, sports team picture, or images of injuries or scars to avoid legal penalties or respond to bad-faith reports.
Civil liberties groups such as the Electronic Frontier Foundation (EFF) have noted that the language of the law itself raises censorship concerns because it’s vague and therefore easily exploited:
Take It Down creates a far broader internet censorship regime than the Digital Millennium Copyright Act (DMCA), which has been widely abused to censor legitimate speech. But at least the DMCA has an anti-abuse provision and protects services from copyright claims should they comply. This bill contains none of those minimal speech protections and essentially greenlights misuse of its takedown regime … Congress should focus on enforcing and improving these existing protections, rather than opting for a broad takedown regime that is bound to be abused. Private platforms can play a part as well, improving reporting and evidence collection systems.
Nor does the law cover the possibility of people filing bad-faith reports.
In the 2002 case Ashcroft v. Free Speech Coalition, the Court said the language of the Child Pornography Protection Act (CPPA) was so broad that it could have been used to censor protected speech. Congress passed the CPPA to combat the circulation of computer-generated child pornography, but as Justice Anthony Kennedy explained in the majority opinion, the language of the CPPA could be used to censor material that seems to depict child pornography without actually doing so.
While we must acknowledge that online exploitation is a very real issue, we cannot solve the problem at the expense of other liberties.
Also in 2002, the Supreme Court heard the case Ashcroft v. ACLU, which came about after Congress passed the Child Online Protection Act (COPA) to prevent minors from accessing adult content online. But again, due to the broad language of the bill, the Court found this law would restrict adults who are within their First Amendment rights to access mature content.
As with the Take It Down Act, here too were laws created to protect children from sexual exploitation online, yet established using vague and overly broad standards that threaten protected speech.
But unfortunately, stories like the one at Aledo High are becoming more common as AI becomes more accessible. Last year, boys at Westfield High School in New Jersey used AI to circulate fake nudes of Francesca Mani, who is 14 years old, and other girls in her class. But Westfield High administrators were caught off guard as they had never experienced this type of incident. Although the Westfield police were notified and the perpetrators were suspended for up to 2 days, parents criticized the school for their weak response.
So to Speak podcast: ‘Robotica: Speech Rights & Artificial Intelligence’
A year later, the school district developed a comprehensive AI policy and amended their bullying policy to cover harassment carried out through “electronic communication” which includes “the use of electronic means to harass, intimidate, or bully including the use of artificial intelligence “AI” technology.” What’s true for Westfield High is true for America — existing laws are often more than adequate to deal with emerging tech issues. By classifying AI material under electronic communication as a category of bullying, Westfield High demonstrates that the creation of new AI policies are redundant. On a national scale, the same can be said for classifying and prosecuting instances of child abuse online.
While we must acknowledge that online exploitation is a very real issue, we cannot solve the problem at the expense of other liberties. Once we grant the government the power to silence the voices we find distasteful, we open the door to censorship. Though it is essential to address the very real harms of emerging AI technology, we must also keep our First Amendment rights intact.
AI speech is speech, and the government shouldn’t get to rewrite it. But across the country, officials are pressuring AI developers to bend outputs to their political preferences.
That danger isn’t theoretical. In July, Missouri’s (now former) Attorney General Andrew Bailey sent OpenAI a letter threatening to investigate the company. In it, Bailey accused their AI chatbot ChatGPT of partisan bias after it ranked President Donald Trump lowest among recent presidents on anti-Semitism. Calling the answer “objectively” wrong, Bailey’s letter cites Trump’s relocation of the U.S. embassy to Jerusalem, the Abraham Accords, and his Jewish family ties as proof the ranking defies “objective facts.”
Although no lawsuit was filed, the looming threat no doubt put considerable pressure on the company to revise its outputs — a preview into how common and far-reaching such tactics will become if courts ever say, as some critics of AI are arguing, that AI speech isn’t explicitly protected by the Constitution.
Lawsuits against Character.AI — another chatbot geared more towards companionship and casual conversation — such as Garcia v. Character Technologies, Inc., show that judges are already being asked to decide whether AI outputs are speech or something else entirely. If courts adopt the view that AI isn’t protected by the First Amendment, nothing would stop government officials from just mandating outputs rather than applying pressure. That’s why FIRE filed an amicus curiae “friend-of-the-court” brief in this litigation to emphasize that the First Amendment shields this expressive technology.
Free expression shouldn’t rise and fall with the party in power, forcing AI engineers to reshape their models to fit every new political climate.
The First Amendment’s protections don’t vanish simply because artificial intelligence is involved. AI is another medium (or tool) for expression. The engineers behind it and the users who prompt it are practicing their craft in much the same way writers, directors, and journalists are practicing theirs. So when officials pressure AI developers to alter or delete outputs, they’re censoring their speech.
By framing ChatGPT’s ranking as “consumer misrepresentation,” Bailey tried to turn protected political speech into grounds for a fraud investigation. Instead of using consumer protection laws for their intended purpose — to, for example, investigate faulty toasters or false advertising — Bailey’s gambit bends them into mechanisms for censoring AI-generated speech. The letter signals to every developer that just one politically sensitive answer could yield a government investigation.
The irony here is striking: Bailey represented the state of Missouri in Murthy v. Missouri, the high-profile lawsuit accusing the Biden administration of jawboning social-media platforms into removing COVID-19 content. In that case, Bailey argued the federal government’s nudging violated the First Amendment because it coerced private actors to police speech the government couldn’t ban outright.
Voters want AI political speech protected – and lawmakers should listen
New polling shows voters fear AI — but fear government censorship more. As lawmakers push new rules, are they protecting elections or silencing speech?
Government pressure is already reshaping AI in other ways. OpenAI’s new policy now warns that your ChatGPT conversations may be scanned, reviewed, and possibly reported to the police. This means users are faced with a choice of whether to risk a visit from law enforcement or forgo the benefits these AI tools offer. Absent robust First Amendment safeguards, the result is government censorship (including jawboning) on one side, and surveillance on the other. Both narrow the space for open inquiry that AI ought to expand.
FIRE’s answer is for the government to first apply the First Amendment appropriately to AI speech, and then improve government transparency to ensure the government is doing so. Our Social Media Administrative Reporting Transparency (“SMART”) Act would require federal officials to disclose their communications with an interactive computer service (like a chatbot) about moderating content. This way users, developers, and the public can see when officials try to influence what AI says. Similar state-level reforms could ensure that no government coercion occurs in the dark.
Free expression shouldn’t rise and fall with the party in power, forcing AI engineers to reshape their models to fit every new political climate. If we want AI to widen the marketplace of ideas, strong First Amendment protections are the place to start.
Jason Clare said TEQSA can only act either as a sledgehammer or a feather. Picture: Martin Ollman
The federal government has published a consultation paper calling for suggestions to reform the Tertiary Education Quality and Standards Agency (TEQSA) Act, which determines the regulator’s powers.
Please login below to view content or subscribe now.
This audio is auto-generated. Please let us know if you have feedback.
Dive Brief:
The Trump administration acted illegally when it delayed and canceled billions of dollars of biomedical research grants, despite Congress appropriating funds to the National Institutes of Health, the Government Accountability Office said on Tuesday.
Between February and June, the watchdog agency estimates the NIH awarded about $8 billion less in funds for research grants and awards compared to the prior year and cut more than 1,800 active grants as it attempted to follow President DonaldTrump’s directive to weed out “equity-related” projects.
The GAO’s report concludes the administration violated the Impoundment Control Act, which requires the president to provide notice before delaying or blocking congressionally directed spending. GAO can file a lawsuit in an attempt to restore the grants. However, the watchdog agency has not yetopted to do so in its dealings with Trump.
Dive Insight:
NIH grants came under scrutiny this winter following a series of executive orders directing federal agencies to terminate “equity-related” grants or contracts, federal funding of projects supporting “gender ideology” and DEI programs.
The NIH began carrying out these directives in February. In addition to the grant cuts, the agency also dragged its feet on approving new projects, GAO found. From late January to early March, the NIH paused grant reviews entirely, delaying funds from being allocated to hospitals and universities.
When contacted for comment, a spokesperson for the HHS referred Healthcare Dive to the agency’s testimony to GAO. The testimony states that NIH has since “moved rapidly to reschedule and hold meetings impacted by the short pause, and to process grant applications.”
The HHS said between March 24 and June 30, NIH scheduled or held 837 peer review meetings — 186 more than for the same period the year prior.
Still, GAO said that the department hadn’tadequately explained its decision to pause the review process in the first place, despite its resumption of grant review.
“If the executive branch wishes to make changes to the appropriation provided to NIH, it must propose funds for rescission or otherwise propose legislation to make changes to the law for consideration by Congress,” the watchdog group wrote in its report. The HHS had done neither, the GAO said, adding: “In short, HHS has offered no evidence that it did not withhold amounts from obligation or expenditure, and it has not shown that the delay was a permissible programmatic one.”
The report also suggests the Trump administration may be continuing efforts to block NIH funds from flowing to medical research.
The office said the Office of Management and Budget asked NIH to “pause the issuing of grants, research contracts and training” in late July. There are reports that decision was later reversed, but GAO said it could not confirm whether the pause was lifted.
Following the release of the report, Democratic lawmakers called for the Trump administration to resume funding NIH grants as Congress specified, warning that medical research progress is at stake.
“Cutting off investments Congress has made into research that saves millions of lives is as backward and as inexcusable as it gets,” said Sen. Patty Murray, D-Wash., in a statement. “It is critical President Trump reverse course, stop decimating the NIH, and get every last bit of this funding out.”
This report is not the first time Trump’s funding cuts have been challenged.
Researchers, unions and a coalition of 16 states sued over the NIHcuts, with academics saying they needed the funds to perform critical medical research, including learning about alcohol’s impact on Alzheimer’s risk and suicide prevention among LGBTQ+ youth experiencing homelessness. In June, a U.S. district judge ordered the NIH to reinstate the plaintiffs’ canceled funds. However, litigation remains ongoing after the Trump administration appealed that ruling.
GAO has the potential to bring its own suit against the NIH, but it will likely be a last resort, according to reporting by the New York Times. The watchdog group has previously found the administration violated the ICA on a range of topics and opted not to sue.
Alexandre de Moraes, the polarizing Brazilian Supreme Court Justice, is no friend to free speech. Though he is a popular figure within Brazil among those who see him as a protector of democracy, he has aggressively wielded his authority to censor, especially on the internet, with little transparency.
From his position in Brazil’s Supreme Court, de Moraes has doggedly pursued wide swaths of speech and speakers off and on the internet, as well as the tech companies hosting them. In a highly public incident last year, Brazil blocked X — and even threatened VPN users accessing it with massive fines — over the company’s noncompliance with de Moraes’ orders.
The actions of de Moraes, and Brazil’s Supreme Court more broadly, have repeatedly drawn the ire of the Trump administration. But chief among President Trump’s grievances is the prosecution of his political ally, former Brazilian President Jair Bolsonaro, who is accused of attempting a coup to overturn his 2022 election loss to President Luiz Inácio Lula da Silva.
How has the Trump administration responded?
Last month, the administration enacted a series of punishments against Brazil’s leadership and de Moraes specifically. In a July 30 executive order, Trump announced tariffs and other sanctions due to Brazil’s prosecution of Bolsonaro and other actions that “conflict with and threaten the policy of the United States to promote free speech and free and fair elections at home and abroad.” The order follows Trump’s weeks-earlier threat of tariffs over the “witch hunt” against Bolsonaro.
Secretary of State Marco Rubio also revoked the visas of de Moraes “and his allies on the court” and their families. And under the Global Magnitsky Human Rights Accountability Act, usually reserved for the most serious human rights abuses, the Department of the Treasury announced sanctions targeting any of de Moraes’s U.S. assets.
Unprincipled, partisan free speech advocacy is no free speech advocacy at all
There is plenty to debate about how to best protect free speech on the global internet, and around the world more generally, and what actions the United States can take in its defense. But, even though Brazil’s adversarial relationship with free expression is deeply alarming, it’s impossible to ignore the incongruity of the Trump administration putting itself in the position of diagnosing and treating government censorship.
Physician, heal thyself.
The opening months of Trump’s second term in office have offered a nonstop, headspinning bonanza of violations, threatened and enacted, against Americans’ First Amendment rights.
I write regularly in the Free Speech Dispatch about the myriad threats to freedom of expression, from Russia to the UK to India to Hong Kong. It’s painfully, brutally clear we need leadership to push back against the wave of global repression that threatens all of our rights. But that leadership must practice what it preaches and avoid simply using concerns about free speech as a pretext to fight partisan political battles. On both counts, this administration has failed.
You will make no converts to the free speech cause by proving right the critics who suspect its advocates are guided by partisan aims, not principled ones. Instead, you will breed cynicism and harm the very cause you claim to support.
This same posturing marred Vice President JD Vance’s objections to European censorship, an ugly trend that’s in dire need of principled critiques. Instead, Vance claimed that under Trump, the “new sheriff in town,” the administration “may disagree with your views, but we will fight to defend your right to offer them in the public square.”
The ugly reality is that the U.S. is rapidly ceding its moral authority to criticize foreign governments’ censorship, like that emanating from Brazil’s Supreme Court, when its own president and agencies are gleefully flouting the First Amendment and free speech principles day in and day out.
Perhaps most baffling was the administration’s objection to the Brazilian government’s targeting of Paulo Figueiredo, a Brazilian journalist, and “U.S. resident, for speech he made on U.S. soil.” Readers may also be able to think of some more government officials targetingimmigrants legally residing in the U.S. for protected speech made on U.S. soil — and they’re doing so from our White House and State Department, not thousands of miles away.
Global censorship is a real challenge, and it’s only getting worse. But until the U.S. removes the censorial beam from its own eye, we may find that other nations are unmoved by our criticisms and cures. Or, they may perhaps even be interested in doling them out to us.
For the last 15 years, science teacher Jeff Grant has used information on climate change from the federal website Climate.gov to create lesson plans, prepare students for Advanced Placement tests and educate fellow teachers. Now, Grant says, he is “grabbing what [he] can” from the site run by the National Oceanic and Atmospheric Administration’s Climate Program Office, amid concerns that the Trump administration is mothballing it as part of a broader effort to undermine climate science and education.
“It’s just one more thing stifling science education,” said Grant, who teaches at Downers Grove North High School in the Chicago suburbs.
Since early May, all 10 editorial contributors to Climate.gov have lost their jobs, and the organization that produces its education resources will soon run out of money. On June 24, the site’s homepage was redirected to NOAA.gov, a change NOAA said was made to comply with an earlier executive order on “restoring gold standard science.” Those steps follow many others the president has made to dismantle federal efforts to fight climate change, which his administration refers to as the “new green scam.”
Former employees of Climate.gov and other educators say they fear that the site, which will no longer produce new content, could be transformed into a platform for disinformation.
“It will make it harder for teachers to do a good job in educating their students about climate change,” said Glenn Branch, deputy director of the nonprofit National Center for Science Education. “Previously, they could rely on the federal government to provide free, up-to-date, accurate resources on climate change that were aimed at helping educators in particular, and they won’t be able to do so if some of these more dire predictions come to pass.”
Such concerns have some foundation. For example, Covid.gov, which during the Biden administration offered health information and access to Covid-19 tests, has been revamped to promote the controversial theory that the coronavirus was created in a lab. The administration has also moved aggressively to delete from government sites other terms that are currently out of favor, such as references to transgender people that were once on the National Park Service website of the Stonewall National Memorial, honoring a major milestone in the fight for LGBTQ+ rights.
Kim Doster, director of NOAA’s office of communications, declined to answer specific questions but shared a version of the statement posted on the NOAA website when Climate.gov was transferred. “In compliance with Executive Order 14303, Restoring Gold Standard Science, NOAA is relocating all research products from Climate.gov to NOAA.gov in an effort to centralize and consolidate resources,” it says.
Related: Want to read more about how climate change is shaping education? Subscribe to our free newsletter.
Climate.gov, founded in 2010 to support earth science instruction in schools, had become a go-to site for educators and the general public for news and information about temperature, sea level rise and other indicators of global warming.
For many educators, it has served a particularly key role. Because its resources are free, they are vital in schools that lack resources and funding, teachers and experts say.
Rebecca Lindsey, Climate.gov’s lead editor and writer, was one of several hundred NOAA probationary employees fired in February, then rehired and put on administrative leave, before being terminated again in March. The rest of the content production team — which included a meteorologist, a graphic artist and data visualizers — lost their jobs in mid-May. Only the site’s two web developers still have their jobs.
A screenshot of the Sea Level Rise Viewer, an interactive NOAA that’s listed as a resource on Climate.gov, a government climate website. Credit: NOAA Office for Coastal Management
Lindsey said she worries that the government “intended to keep the site up and use it to spread climate misinformation, because they were keeping the web developers and getting rid of the content team.”
In addition, the Climate Literacy and Energy Awareness Network, the official content provider for the education section of the site, has not received the latest installment of its three-year grant and expects its funds to run out in August.
“We won’t have funding to provide updates, fix hyperlinks and make sure that new resources are being added, or help teachers manage or address or use the resources,” said Anne Gold, CLEAN’s principal investigator. “It’s going to start deteriorating in quality.”
CLEAN, whose website is hosted by Carleton College, is now searching for other sources of money to continue its work, Gold said.
With the June 24 change redirecting visitors from Climate.gov to NOAA.gov/climate, the website for the first time falls under the purview of a political appointee: Doster. Its previous leader, David Herring, is a science writer and educator.
Melissa Lau, an AP environmental science teacher in Piedmont, Oklahoma, said the relocated site was “really difficult to navigate.”
As someone who lives in Tornado Alley, Lau said, she frequented CLEAN and NOAA sites to show her students localized, real-time data on storm seasons. She said she is concerned that teachers won’t have time to track down information that was shifted in the website’s move and, as a result, may opt not to teach climate change.
The executive order on “restoring gold standard science” that appears to have triggered the shift gives political appointees the authority to decide what science information needs to be modified to align with its tenets.
While the disclaimer posted to NOAA.gov seems to imply that Climate.gov did not meet this requirement, educators and researchers said that the site and its CLEAN education resources were the epitome of a gold standard.
“I want to stress that the reason why CLEAN is considered the gold standard is because we have such high standards for scientific accuracy, classroom readiness and maintenance,” Gold said. “We all know that knowledge is power, and power gives hope. … [Losing funding] is going to be a huge loss to classrooms and to students and the next generation.”
This is only the latest attack by the Trump administration on education around climate change. This month, the U.S. Global Change Research Program’s website, GlobalChange.gov, was shut down by the administration, after the program was defunded in April. The website once hosted an extensive climate literacy guide, along with all five iterations of the National Climate Assessment — a congressionally required report that informed the public about the effects and risks of climate change, along with local, actionable responses.
The Department of Commerce, which oversees NOAA, has cut other federal funding for climate research, including at Princeton University, arguing that these climate grant awards promoted “exaggerated and implausible climate threats, contributing to a phenomenon known as ‘climate anxiety,’ which has increased significantly among America’s youth.”
“The more you know [about climate change], the more it’s not a scary monster in the closet,” said Lauren Madden, professor of elementary science education at the College of New Jersey. “It’s a thing you can react to.” She added, “We’re going to have more storms, we’re going to have more fires, we’re going to have more droughts. There are things we can do to help slow this. … I think that quells anxiety, that doesn’t spark it.”
And climate education has broad public support — about 3 in 4 registered voters say schools should teach children about global warming, according to a 2024 report from the Yale Program on Climate Change Communication. Similarly, 77 percent of Americans regard it as very or somewhat important for elementary and secondary school students to learn about climate change, according to a 2019 study. And all but five states have adopted science standards that incorporate at least some instruction on climate change.
As a result, many science teachers rely on federal tools and embed them in their curriculum. They are worried that the information will no longer be relevant, or disappear entirely, according to Lori Henrickson, former climate integration specialist for Washington state’s education department. Henrickson, who lost her job this June as the result of state budget cuts, was in charge of integrating climate education across content areas in the state, from language arts to physical education.
The .gov top-level domain connotes credibility and accessibility, according to Branch: “It is also easier for teachers facing or fearing climate change denial backlash to cite a reliable, free source from the federal government.”
With Climate.gov’s future uncertain, educators are looking to other resources, like university websites and tools from other countries.
“I’m sure there will continue to be tools, and there will be enough people who will be willing to pay to access them,” Madden said. But, she added, “they probably won’t be as comprehensive, and it won’t feel like it’s a democratic process. It’ll feel like: If you or your employer are willing to chip in for it, then you’ll have access.”
“I feel like with all the federal websites, I’m constantly checking to see what’s still up and what’s not,” Madden said.
Bertha Vazquez, education director for the Center of Inquiry, an organization that works to preserve science and critical thinking, said she worried that the disappearance of climate information could leave U.S. students behind.
“The future of the American economy is not in oil, the future of the American economy is in solar and wind and geothermal. And if we’re going to keep up with the international economy, we need to go in that direction,” she said. But while the U.S. should be leading the way in scientific discovery, Vazquez said, such work will now be left to other countries.
Lau said she felt helpless and frustrated about Climate.gov’s shutdown and about the “attack on American science in general.”
“I don’t know what to do. I can contact my legislators, but my legislators from my state are not going to be really open to my concerns,” she said. “If students next year are asking me questions about [science research and funding], I have to tell them, ‘I do not know,’ and just have to leave it at that.”
Contact editor Caroline Preston at 212-870-8965, via Signal at CarolineP.83 or on email at [email protected].
The Hechinger Report provides in-depth, fact-based, unbiased reporting on education that is free to all readers. But that doesn’t mean it’s free to produce. Our work keeps educators and the public informed about pressing issues at schools and on campuses throughout the country. We tell the whole story, even when the details are inconvenient. Help us keep doing that.
When the franchising scandal first broke, many thought it was going to be a flash in the pan, an airing of the darkest depths of the sector but something that didn’t really impact the mainstream.
That hasn’t been the case.
The more it digs, the more concerned the government seems to get, and the proposed reforms to register the largest delivery partners seem unlikely to mark the end of its attention.
Last orders
The sector would be foolish to wait for the Government’s response to its consultation, or for the Office for Students to come knocking. Subcontracted provision in England has increased 358 per cent over the past five years: and, for some providers this provision significantly outnumbers the students they teach directly themselves. Franchised business and management provision has grown by 44 per cent, and the number of students from IMD quintile 1 (the most deprived) taught via these arrangements have increased 31 per cent, compared to an overall rise in student numbers of 15 per cent.
The sector talks a big game about institutional autonomy – and they’re right to do so; it is a vital attribute of the UK sector. But it shouldn’t be taken for granted, and that means demonstrating clear action when practices are scrutinised.
Front foot
So today, QAA has released new comprehensive guidance (part of a suite sitting underneath the UK Quality Code) to help the sector get on the front foot. For the first time since the franchising scandal broke, experts from across the UK sector have developed a toolkit for anyone working in partnerships to know what good practice can look like, what questions they should be asking themselves, and how their own provision stacks up against what others are doing.
The guidance is framed around three discrete principles: all partnerships should add direct value to the staff and student experience and widen learning opportunities; academic standards and the quality of the student experience should not be compromised; and oversight should be as rigorous, secure and open to scrutiny as the provision delivered by a single provider. All partners share responsibility for the student learning experience and the academic standards students are held to, but it is the awarding partner who is ultimately accountable for awards offered in its name.
If you’re working in partnership management and are concerned about how your institution should be responding to the increased scrutiny coming from government, the guidance talks you through each stage of the partnership lifecycle, with reflective questions and scenarios to prompt consideration of your own practice. And as providers put the guidance and its recommendations into practice, they will be able to tell a more convincing and reassuring story about how they work with their partners to deliver a high quality experience.
Starter for five
But the sector getting its house in order will only quell concerns if those scrutinising feel assured of provider action. So for anyone concerned, we’ve distilled five starter questions from the guidance that we’d expect any provider to be able to answer about their partnerships.
Are there clear and shared academic standards? Providers should be able to provide agreed terms on academic standards and quality assurance and plans for continuous improvement.
Is oversight tailored to risk? Providers who have a large portfolio should be able to demonstrate how they take an agile, proportionate approach to each partnership.
What are the formal governance and accountability mechanisms? A provider’s governors or board should be able to tell you what decisions have been made and why.
How is data used to drive performance and mitigate risk? Providers should be able to tell you what data they have and what it tells them about their partnerships and the students’ experience, and any actions they plan to take.
And finally, how does your relationship enable challenge and improvement? Providers should be able to tell you when they last spoke to each of their partners, what topics were discussed and lead providers should be able to detail what mechanisms they use to hold their partners to account when issues arise.
Integrity and responsibility
The government has a duty to prevent misuse of public money and to ensure the integrity of a system that receives significant amounts of it. The regulator has a responsibility to investigate where it suspects there is poor practice and to act accordingly. But the sector has a responsibility – both to its students and, also, to itself – to respond to the legitimate concerns raised around partnership provision and to demonstrate it’s taking action. This lever is just as, if not more, important, because government and regulatory action becomes more necessary and more stringent if we don’t get this right.
The sector cannot afford not to grasp the nettle on this. Public trust, the sector’s reputation and, most importantly, the learning experience students deserve, are all on the line.
QAA’s guidance is practical, expert-informed and rooted in shared principles to help providers not only meet expectations but lead the way in restoring confidence. Because if the sector doesn’t demonstrate its commitment to action on this, the government and the regulator surely will.
HEPI Director, Nick Hillman, looks at the latest row on admissions to the University of Oxford.
In a speech on Friday, the Minister for Skills, Baroness Smith, strongly chastised her alma mater, the University of Oxford, for taking a third of their entrants from the 6% of kids that go to private schools.
In a section of the speech entitled ‘Challenging Oxford’, we were told the situation is ‘absurd’, ‘arcane’ and ‘can’t continue’:
Oxford recently released their state school admissions data for 2024.
And the results were poor.
66.2% – the lowest entry rate since 2019.
I want to be clear, speaking at an Oxford college today, that this is unacceptable.
The university must do better.
The independent sector educates around 6% of school children in the UK.
But they make-up 33.8% of Oxford entrants.
Do you really think you’re finding the cream of the crop, if a third of your students come from 6% of the population?
It’s absurd.
Arcane, even.
And it can’t continue.
It’s because I care about Oxford and I understand the difference that it can make to people’s lives that I’m challenging you to do better. But it certainly isn’t only Oxford that has much further to go in ensuring access.
This language reminded me of the Laura Spence affair, which produced so much heat and so little light in the Blair / Brown years and which may even have set back sensible conversations on broadening access to selective higher education.
I wrote in a blog over the weekend that the Government are at risk of forgetting the benefit of education for education’s sake. That represents a political hole that Ministers should do everything to avoid as it could come to define them. Ill-thought through attacks on the most elite universities for their finely-grained admissions decisions represent a similar hole best avoided. Just imagine if the Minister had set out plans to tackle a really big access problem, like boys’ educational underachievement, instead. The Trump/Harvard spat is something any progressive government should seek to avoid, not copy.
The latest chastisement is poorly formed for at least three specific reasons: the 6% is wrong in this context; the 33.8% number does not tell us what people tend to think it does; and Oxford’s current position of not closely monitoring the state/independent split is actually in line with the regulator’s guidance.
6% represents only half the proportion (12%) of school leavers educated at independent schools. In other words, the 6% number is a snapshot for the proportion of all young people in private schools right now; it tells us nothing about those at the end of their schooling and on the cusp of higher education.
The 33.8% number is unhelpful because 20%+ of Oxford’s new undergraduates hail from overseas and they are entirely ignored in the calculation. If you include the (over) one in five Oxford undergraduate entrants educated overseas, the proportion of Oxford’s intake that is made up of UK private school kids falls from from something like one-third to more like one-quarter. This matters in part because the number of international students at Oxford has grown, meaning there are fewer places for home students of all backgrounds. In 2024, Oxford admitted 100 more undergraduate students than in 2006, but there were 250 more international students – and consequently fewer Brits. We seem to be obsessed with the backgrounds of home students and, because we want their money, entirely uninterested in the backgrounds of international students.
The Office for Students dislikes the state/private metric. This is because of the differences within these two categories: in other words, there are high-performing state schools and less high-performing independent schools. Last year, when the University of Cambridge said they planned to move away from a simplistic state/independent school target, John Blake, the Director of Fair Access and Participation at the Office for Students, confirmed to the BBC, ‘we do not require a target on the proportion of pupils from state schools entering a particular university.’ So universities have typically shied away from this measure in recent times. If Ministers think it is a key metric after all and if they really do wish to condemn individual institutions for their state/independent split, it would have made sense to have had a conversation with the Office for Students and to have encouraged them to put out new guidance first. At the moment, the Minister and the regulator are saying different things on an important issue of high media attention.
Are independently educated pupils overrepresented at Oxbridge? Quite possibly, but the Minister’s stick/schtick, while at one with the Government’s wider negative approach to independent schools, seems a sub-optimal way to engineer a conversation on the issue. Perhaps Whitehall wanted a headline more than it wanted to get under the skin of the issue?
we do not require a target on the proportion of pupils from state schools entering a particular university
John Blake, Director for Fair access and participation
WASHINGTON, D.C. — As the House Select Committee on the China opens its landmark hearing, “Authoritarians and Algorithms: Why U.S. AI Must Lead,” Committee leaders are unveiling new bipartisan legislation to confront the CCP’s growing exploitation of artificial intelligence.
Chairman John Moolenaar (R-MI) and Ranking Member Raja Krishnamoorthi (D-IL) today announced the House introduction of the “No Adversarial AI Act”—bipartisan legislation also being championed in the Senate by Senators Rick Scott (R-FL) and Gary Peters (D-MI). The bill would prohibit U.S. executive agencies from acquiring or using artificial intelligence developed by companies tied to foreign adversaries like the Chinese Communist Party. The House legislation is cosponsored by a bipartisan group of Select Committee members, including Reps. Ritchie Torres (D-NY) and Darin LaHood (R-IL).
“We are in a new Cold War—and AI is the strategic technology at the center,” said Chairman Moolenaar.“The CCP doesn’t innovate—it steals, scales, and subverts. From IP theft and chip smuggling to embedding AI in surveillance and military platforms, the Chinese Communist Party is racing to weaponize this technology. We must draw a clear line: U.S. government systems cannot be powered by tools built to serve authoritarian interests.”
What the No Adversarial AI Act Does:
Creates a public list of AI systems developed by foreign adversaries, maintained and updated by the Federal Acquisition Security Council.
Prohibits executive agencies from acquiring or using adversary-developed AI—except in narrow cases such as research, counterterrorism, or mission-critical needs.
Establishes a delisting process for companies that can demonstrate they are free from foreign adversary control or influence.
“Artificial intelligence controlled by foreign adversaries poses a direct threat to our national security, our data, and our government operations,” said Ranking Member Raja Krishnamoorthi. “We cannot allow hostile regimes to embed their code in our most sensitive systems. This bipartisan legislation will create a clear firewall between foreign adversary AI and the U.S. government, protecting our institutions and the American people. Chinese, Russian, and other adversary AI systems simply do not belong on government devices, and certainly shouldn’t be entrusted with government data.”
Senator Rick Scott said, “The Communist Chinese regime will use any means necessary to spy, steal, and undermine the United States, and as AI technology advances, we must do more to protect our national security and stop adversarial regimes from using technology against us. With clear evidence that China can have access to U.S. user data on AI systems, it’s absolutely insane for our own federal agencies to be using these dangerous platforms and subject our government to Beijing’s control. Our No Adversarial AI Act will stop this direct threat to our national security and keep the American government’s sensitive data out of enemy hands.”
The legislation marks a major action in the Select Committee’s AI campaign, which aims to secure U.S. AI supply chains, enforce robust export controls, and ensure American innovation does not fuel authoritarian surveillance or military systems abroad.
Today’s hearing and legislation continues the series of new proposals and messaging the Committee will roll out this summer to confront the CCP’s exploitation of U.S. innovation and prevent American technology from fueling Beijing’s AI ambitions.
As artificial intelligence technologies make their way into political ads and campaigning, Americans are expressing growing concern. But they’re not just worried about deepfakes and deceptive content’s impact on elections — they also fear how the government might use the fight against misinformation to restrict free speech.
In a recently released FIRE poll of registered American voters, conducted by Morning Consult, one concern stood out: government regulation itself. Nearly half of respondents (45%) said they are “extremely” or “very” concerned that government regulation of election-related AI content could be abused to suppress criticism of elected officials. That’s a powerful signal that while Americans see the risks posed by AI, they don’t trust government regulators to police political expression fairly.
When asked to choose between protecting free speech in politics or stopping deceptive content, a plurality (47%) said protecting free speech in politics is more important, even if that means allowing some deceptive content. Just 37% prioritized stopping deceptive content, even at the expense of limiting speech that would otherwise be protected by the First Amendment. These sentiments are held across the political spectrum, but are stronger among Independents and Republicans, than among Democrats.
This isn’t just a preference — it’s a principled stand in favor of the core freedoms the First Amendment exists to protect. Political speech lies at the heart of those freedoms, and Americans clearly recognize that any government attempts to police what can or can’t be said pose a far greater threat to democracy than free speech itself.
Regulation threatens participation
The chilling effects are already measurable. About 28% of voters said they’d be less likely to share content on social media if the government began regulating AI-generated or AI-altered content. (That’s right: All content, not just AI-generated or AI-altered content.) That may not sound dramatic at first glance, but that’s more than the average voter turnout during the last midterm primaries. As our political culture is increasingly shaped online, discouraging speech — even unintentionally — can have real consequences for public discourse.
These findings suggest a troubling trajectory: Government regulations justified in the name of protecting the public from AI could end up silencing the public instead.
While some polls show that a similar percentage of voters (41%) say it’s important to protect people from misinformation, that concern cannot be used to justify censorship. About 39% said that preserving freedom of speech should be the government’s top priority when crafting AI laws. Only 12% said that view doesn’t describe them at all. In other words, most Americans believe that protecting speech isn’t just one goal among many — it’s the central concern.
And they’re right to think so. The First Amendment doesn’t permit the government to restrict speech simply because it believes the public might be misled. The solution to bad speech is more speech, not less.
These results should serve as a warning to policymakers: The public views efforts to regulate AI in political campaigns as a risk to free expression. FIRE has been actively engaged in legislative advocacy to safeguard First Amendment rights, including vague and overbroad bans or disclosure requirements imposed on AI content.
If voters already believe regulation will be abused — and are already pulling back from political expression using AI — that’s not just a theoretical harm. It’s a chilling effect in action.
Instead of rushing to regulate, elected officials should reaffirm their commitment to protecting political speech, no matter the medium. The technology may be new, but the principle is not: In a free society, the government doesn’t get to decide which ideas are too dangerous to be heard.
The poll was conducted May 13-15, 2025, among a sample of registered voters in the US. A total of 2,005 interviews were conducted online across the US for a margin of error of plus or minus 2 percentage points. Frequency counts may not sum to 2,005 due to weighting and rounding. Topline results are available here.
