This audio is auto-generated. Please let us know if you have feedback.
Dive Brief:
Diversity, equity, inclusion and accessibility best practices are not illegal, said Massachusetts Attorney General Andrea Joy Campbell and Illinois AG Kwame Raoul, in a multi-state DEIA at work guidance.
In the Feb. 13 letter, the AGs said the federal government lacks the power to issue executive orders that prohibit “otherwise lawful activities in the private sector or mandates the wholesale removal of these policies and practices within private organizations, including those that receive federal contracts and grants.”
The AGs of Arizona, California, Connecticut, Delaware, Hawaii, Maine, Maryland, Minnesota, Nevada, New Jersey, New York, Oregon, Rhode Island and Vermont joined in issuing the guidance.
Dive Insight:
The letter came as a response to constituent concerns about the continued viability of DEIA, the AGs said, mainly in light of President Donald Trump’s executive orders.
The primary EO in question, “Ending Illegal Discrimination and Restoring Merit-Based Opportunity,” includes a directive that “order[s] all agencies to enforce our longstanding civil-rights laws and to combat illegal private-sector DEI preferences, mandates, policies, programs, and activities.”
The executive order alleges that colleges, along with other organizations, have “adopted and actively use dangerous, demeaning, and immoral race- and sex-based preferences under the guise of so-called … ‘diversity, equity, inclusion, and accessibility.’”
Campbell and Raoul said the order “conflates unlawful preferences in hiring and promotion with sound and lawful best practices for promoting diversity, equity, inclusion, and accessibility in the workforce.”
While the judge did not prevent the U.S. Department of Justice from proceeding with its investigation of private-sector DEI programs, Judge Adam Abelson held that the plaintiffs would likely succeed with their First and Fifth amendment claims, as well as claims alleging violations of the separation of powers clause.
Prior to the most recent guidance, Democrat attorney generals have made it their priority to speak up about DEI: Last summer, the AGs defended the American Bar Association’s diversity requirements for law schools.
More recently, the Democrat AGs said that the U.S. is “on the brink of dictatorship” due to Trump’s executive orders challenging the scope of the Constitution.
A key takeaway for HR? “Properly developed and implemented initiatives aimed at ensuring that diverse perspectives are included in the workplace help prevent unlawful discrimination,” the AGs said.
If you’ve seen the award-winning film “Conclave”, you now know how exciting it can be when Catholicism’s cardinals gather behind closed doors to elect a new pope. The declining health in Pope Francis, 88, means another conclave is coming sooner rather than later.
So who is likely to win? The truth is — we have no idea.
What are the issues the world’s 1.3 billion Catholics would like the next head of their worldwide Church to tackle? We don’t know that for certain either.
In his 12 years as pope, Francis has so scrambled the Church and the traditional paths toward becoming its leader that the conclave — already the strangest election you’ll never see — is even harder than ever to predict.
With the 120 cardinal electors so unsure, unforeseen events like a stirring speech behind closed doors can produce big changes. That’s how Buenos Aires Cardinal Jorge Bergoglio became Pope Francis back in 2013.
Don’t trust the early bets.
Another lesson from recent conclaves is to be very wary of any lists of leading candidates. They are not based on opinion polls or popularity contests like forecasts before political elections.
They are in fact little more than educated guesses by journalists and bookmakers, and can be laughingly far off the mark.
Many lists prior to the 2005 conclave named Milan Cardinal Dionigi Tettamanzi as a leading candidate; he reportedly got only two votes. In 2013, the pro-conclave lists of papabili — potential popes — did not even mention the eventual winner.
But readers always want to know what will happen, and the final result we can’t predict. Since this will be the third conclave I’ve either covered or commented on, let me at least say what to expect.
First of all, if you want, go see the film “Conclave” — at best before we learn if it has won any Oscars. Sure, the film is tenser and more action-packed than a real conclave, and its ending seems improbable. Due to timing limits, the leading characters are painted with a rather cartoonish brush. There are some small mistakes.
But this is entertainment, not a documentary. It is beautifully filmed. It gives an idea of the predictable ritual and possible mishaps that could influence the outcome. It’s worth seeing even if not totally believing.
Politicking for popedom
As the film shows, hopeful candidates drum up support without publicly declaring their candidacy. Open campaigning is out but supporters eagerly swap information about favorites and opponents.
Speeches about the Church during the closed-door “general congregations” meeting the week before the conclave become veiled campaign pitches for the ambitious.
Once they enter the Sistine Chapel for the conclave, the cardinals are cut off from the outside world and sworn to secrecy. This lasts for a few days of voting — a two-thirds majority is needed — until white smoke goes up from the chapel’s chimney and the new pope appears in public for the first time on the balcony of St. Peter’s Basilica next door.
What makes this conclave even more unpredictable than earlier ones is that Pope Francis has often overlooked traditional cardinal’s seats in Europe to give red hats to lesser-known and more pastorally-minded bishops from his beloved “peripheries.”
There are now cardinals in unlikely places such as Yangon, Ulaanbaatar, Algiers and Tehran, representing minuscule communities of Catholics. It’s hard to say what these prelates think or how they will vote.
A new pope could mean a new direction.
Pope Francis has appointed about 80 of the current cardinal electors out of 120, so the two-thirds majority needed for election should be there. But since so many of them are not known in Rome, it’s hard to say whether they want to continue his policies or take the Church in a different direction.
After the 2013 conclave, several conservative Catholic groups — mostly in the United States — disapproved of Pope Francis’s more open style. They said cardinals did not have enough information before they voted him in, and vowed to publish detailed profiles of all prelates on the internet.
The College of Cardinals Report seems the furthest advanced, with profiles of 40 cardinals with their positions on key issues like abortion or woman priests. It is headed by Edward Pentin, a conservative Vatican watcher.
Another project, the Red Hat Report, began in 2018 with lots of publicity saying ex-FBI agents would do some of the research and freelancers would help edit cardinals’ Wikipedia pages. It also leans conservative but has not made much noise recently.
Their main candidate appears to be Budapest Cardinal Péter Erdő, who headed the Council of the Bishops’ Conferences of Europe from 2006 to 2016.
Considered close to Hungarian Prime Minister Viktor Orbán, he seems not to share Pope Francis’s very welcoming views on migration but has hosted two visits by Pope Francis to Budapest.
Power in the Global South
Another conservative noticed is Kinshasa Cardinal Fridolin Ambongo Besungu, although the very traditional stances African prelates take turn off other cardinals.
Progressives mentioned include Bologna Cardinal Matteo Zuppi, head of the Italian Bishops Conference, and Curia Cardinal Luis Tagle, a Filipino once dubbed the “Asian Francis.” But it’s unclear whether a majority of cardinals wants an extension of the Francis years.
There are also moderates such as two Italians — possibly too diplomatic Cardinal Pietro Parolin, the current Secretary of Stage (number two man at the Vatican), and Cardinal Pierbattista Pizzaballa, the Latin Patriarch of Jerusalem — as well as French Cardinal Jean-Marc Aveline, who shares Pope Francis’s interest in the Mediterranean and caring for its migrants.
But with two-thirds of all Catholics now in the Global South, will the conclave return to the Italian and then European monopoly on the papacy after an Argentinian pope?
There are far more questions than answers. We’ll only start to know the responses after the white smoke rises.
Three questions to consider:
1. Why was Pope Francis hailed as a rule-breaker when he was elected to head the Catholic Church in 2013? 2. How does politics play out in the election of a new pope? 3. If you were part of the next conclave what would you be looking for in the candidates for pope?
A new report from the Common App found major racial disparities in persistence rates for students who enter college pursuing degrees in science, technology, engineering, mathematics or medicine.
Just over half of all college applicants express interest in a STEMM field before entering college—except for Asian American students, 72 percent of whom are interested in STEMM. But while more than half of white and Asian students pursuing STEMM obtain a degree in their chosen field within six years, only one-third of first-generation and Latino students who pursue STEMM, and 28 percent of Black or African American students, persist to earn a degree.
The disparities go beyond race. While 54 percent of continuing-generation STEMM students earn a degree in their chosen field, only 34 percent of first-gen students do so. And 51 percent of STEMM-interested students from above the median household income earn a degree in their field, compared to 38 percent of students from below median income levels.
“Our research finds many more talented STEMM aspirants from underrepresented backgrounds applying for college than completing it,” the report concludes.
The study also found that more female STEMM students switch their degree paths (18 percent) than male students (14 percent), though they complete STEMM degrees at similar rates.
Stanford University backed off a plan, almost four years in the making, to buy the Notre Dame de Namur University campus in nearby Belmont, Calif., the San Francisco Chronicle reported.
“The university arrived at this decision after evaluating many factors, some of which could not be anticipated when Stanford first entered into an option purchase agreement with NDNU almost four years ago,” Stanford officials wrote in a Tuesday statement announcing the decision.
Officials added that as the university was “exploring possible academic uses for a Stanford Belmont campus,” it became clear “that identifying and establishing those uses for a potential Belmont campus will take significantly longer than we initially planned.”
Administrators also seemed to hint at potential financial concerns, as President Donald Trump has sought—unsuccessfully, so far—to cap reimbursements for indirect research costs funded by the National Institutes of Health, which experts have warned will harm research universities.
“The landscape for research universities has changed considerably since Stanford entered into the option purchase agreement with NDNU,” Stanford officials wrote. “These changes are resulting in greater uncertainties and a different set of institutional and financial challenges for Stanford.”
In their own statement, NDNU officials noted the university would continue to seek a buyer and expressed disappointment that the sale had fallen through.
Notre Dame de Namur has sought to sell the Belmont campus near Palo Alto since it shrank its offerings and moved a number of its programs online in 2021 amid financial challenges that pushed it to the brink of closure. Now the private Roman Catholic institution is focused on graduate education and offers a mix of in-person, hybrid and online programs.
Officials had expected the sale of the Belmont campus to provide a financial boon.
“Our focus remains on finding a buyer who will preserve and honor the historical significance of this beautiful campus and continue to serve the community-oriented mission that has long been a cornerstone of Notre Dame de Namur University,” NDNU president Beth Martin wrote.
Local lawmakers walked out of a meeting with University of Pennsylvania officials on Tuesday due to what they said was insufficient support for diversity, equity and inclusion, WHYY reported.
Pennsylvania state senator Art Haywood and state representative Napoleon Nelson, both Democrats, reportedly walked out of the meeting after a Penn official referred to diversity as a “lightning rod.”
The meeting, which included several elected state and city officials, became contentious, with lawmakers pressing Penn to hold its ground against the Trump administration’s executive actions on DEI, according to WHYY.
Penn has since removed webpages about its DEI initiatives and updated its nondiscrimination policies, despite swirling legal questions and a nationwide injunction handed down last week that blocked the Trump administration’s plans to crack down on college DEI efforts.
University officials denied backtracking on Penn’s commitment to DEI, according to lawmakers’ accounts of the meeting.
A university spokesperson told the Philadelphia radio station that Penn remains “committed to nondiscrimination in all of our operations and policies” and said the institution appreciated the concerns raised.
Lawmakers indicated that they would continue to press Penn on its commitment to DEI; several provided fiery statements to WHYY casting the university’s response as weak.
“Penn has made a cowardly move, rushing to heed dog-whistle demands from a feckless federal leadership and dismantle their programs that welcome students and workers from an expansive range of backgrounds,” state senator Nikil Saval, a Democrat, told the radio station.
In a time of skyrocketing paper and postage costs, alumni magazines are paradoxically enjoying a renaissance. After cutting back—or cutting down—print issues during the pandemic, many institutions are now pushing for expanded page counts, more copies, better photography, multimedia extras and more institutional support.
Why?
Because audiences appreciate the thought-provoking content and the tangible, premium reminder of the enduring connection with their alma mater. In a 2024 CASE readership survey, 68 percent of TCU Magazine’s readers reported spending 30 minutes or more with every issue. Almost half reported that the magazine was a go-to source for continuing education.
Journalists are pouring their passion and experience into institutional magazines because higher education shines glimmers of hope into an increasingly dark world. They highlight purpose-driven students who will tackle the problems of the future and brilliant faculty whose research is providing innovative solutions to the planet’s most pressing challenges.
Our readership analytics at TCU Magazine have long shown a strong audience appetite for well-researched and carefully written and edited feature stories about forward momentum and its relationship to education. Since 2015, our overall page views have experienced an astounding 1,300 percent growth. That number sounds outlandish, but I can assure you it is accurate.
Our alumni, parents, donors and internal stakeholders are and always have been the primary audiences. But they aren’t the only people who want to know about the students, faculty, staff and initiatives that thrive on our campus. TCU Magazine’s stories are crafted to be relevant far beyond our campus community and long after the initial date of publication.
In 2021, when all the rules were being rewritten, we proposed a partnership with our colleagues in marketing. We suggested a trial run of using existing magazine stories as peer marketing material, promoting those features to internet users who live in the proximity of the country’s top 150 colleges and universities. The goal was for other professionals in higher education to learn about TCU beyond our exceptional student experience and athletic success.
TCU’s marketing director agreed that long-form content could run alongside more traditional digital marketing materials. Why not? Serving stories about improving teacher retirement plans; developing free, open-source digital mapping tools; or better understanding mutations in the BRCA gene benefit us and all manner of readers.
Audiences learn something new and interesting about how research is shaping the future, and we achieve our goal of enhancing TCU’s academic reputation.
Win-win.
Together, we built a partnership with a digital marketing agency based in Fort Worth. With their expert guidance, we got a crash course in the differences between Google Display Network and SEM keywords, Demand Gen ad placements, bidding strategies, and the wisdom of narrowing ad placements in social media feeds.
We launched our first joint academic content campaign in April 2021 with a modest investment. The results were promising: In two months, we got the TCU initials in front of more than six million people around the country and enticed 87,000 of those people to click on the ad and come to the website to read the story.
Best of all, these were what we refer to as quality clicks, because the average reader spent almost two minutes on one of our stories, far above the internet’s long-form content average of less than 40 seconds. That small trial convinced our divisional leaders that magazine material could be marketing gold.
We didn’t need to reinvent the wheel or invest in outside development of marketing-specific content because we had a treasure trove already flowing from a steady creative stream inside our office.
We expanded the efforts in 2022, sharing new stories with 10.5 million pairs of eyes and bringing 116,000 more people to our site to learn about TCU research. That year, we got an email from Puerto Rico about French professor Benjamin Ireland’s research reuniting families torn apart during forced internment during World War II. “I am not sure why Facebook ‘promoted’ your article to me this morning,” the effusive author shared, “but something made me click to read more.”
We’ve continued to grow these campaigns. Though our mission at the magazine is and always will be to serve the TCU community first, we now factor in whether a proposed story might have a broader impact or might help us tell a more expansive tale about how the type of ethical leadership that flourishes here and makes the world a better place.
My opinion is that these campaigns have worked because they’re a perfect merger of marketing and communication. We’re doing what magazine writers and editors have always done—telling authentic stories about real people doing purpose-driven work.
What’s not to like?
Caroline Collier is director of editorial services at Texas Christian University and editor of TCU Magazine.
Lately, I have been experiencing anger, occasionally edging toward rage (depending on my mood) when I open a new document in MSWord and I see the ghostly prompt urging me to use its Copilot generative AI tool.
I do not want to use this tool. I especially do not want to use this tool to start a draft of a document, because writing the first draft under the power of my own thoughts is the key to ultimately producing something someone else might want to read, and outcome on which my living depends, but it’s also, the point of all writing ever, in any context, as far as I’m concerned.
I am persuaded by Marc Watkins’s framing of “AI is unavoidable, not inevitable” for no other reason than the tech companies will not allow us to avoid their generative AI offerings. We can’t get away from this stuff if we want to, and boy, do I really want to.
But just because it is unavoidable and must be acknowledged and, in its way, dealt with, does not mean we are required to use or experiment with it. Over the period of writing More Than Words: How to Think About Writing in the Age of AI, and now spending a month or so promoting and talking about the book in various venues, I grow more and more convinced that if this technology is to have utility in helping students learn—and I mean learn, not merely do school—this utility is likely to be specialized and narrow and the product of deep thought and careful exploration and step-by-step iteration.
Instead, we’re on the receiving end of a fire hose spraying, This is the future!
Is it, really?
One of the reasons we’re being told it’s the future is because at this time, generative AI has no strong business rationale. Don’t take my word for it. Listen to Microsoft CEO Satya Nadella, who admitted in a podcast interview that generative AI applications have had no meaningful effect on GDP, suggesting they are not amazing engines of increased productivity.
Tech watcher Ed Zitron has been saying for months that there is no “AI revolution” and that we’re heading toward the bursting of a bubble that will at least rival the 2008 downturn caused by the subprime mortgage crisis.
So, while there is reason to believe that we are experiencing a bubble that is inevitably going to burst, as we imagine what our institutional and individual relationships should be with this technology, I think it’s useful to see what the people who are—literally—invested in AI envision for our futures. If they are right, and AI is inevitable, what awaits us?
Let’s check in with the people directly funding and developing AI technology what they foresee for the educators of the United States.
@elonmusk/X
That is the man who is apparently running—and running roughshod over—the United States government suggesting that AI-assisted education is superior to what teachers deliver. Now, we know this is not true. We know it will never be true—that is, unless what counts as outcomes is defined down to what AI-assisted education can deliver.
“But to be clear, the ‘better outcomes’ that Silicon Valley shit-posters Palmer Luckey and Elon Musk fantasize about in the image above do not involve the quality of education—of learning or teaching or schooling. (You’re not fooled that they do, right?) They aren’t talking about improved test scores or stronger college admissions or nicer job prospects for graduates or well-compensated teachers or happier, healthier kids or any such metric. Rather, this is a call for AI to facilitate the destruction of the teaching profession, one that is, at the K-12 level comprised predominantly of women (and, in the U.S., is the largest union) and at the university level—in their imaginations, at least—is comprised predominantly of ‘woke.’”
It is hard to know what to do about a technology that some intend to leverage to destroy your profession and harm the constituents your profession is meant to serve. More Than Words is not a book that argues we must resist this technology at all costs, but again, these people want to destroy me, you, us.
ChatGPT and its ilk haven’t even been around for all that long, and we already see the consequences of voluntary deskilling. Futurism reports, “Young coders are using AI for everything, giving ‘blank stares’ when asked how programs actually work.”
Namanyay Goel, a veteran coder who has been observing the AI-wielding coders who can’t actually code, says, “The foundational knowledge that used to come from struggling through problems is just … missing.” This is output divorced from process, a pattern that is already endemic to our transactional model of schooling, but which AI now supercharges.
There is no role for educational institutions in the world where we allow this sort of thing to substitute for knowledge and learning. That may be the least of our problems should the full deskilling result. (See the film Idiocracy for that particular flavor of dystopia.)
When Microsoft shoves its AI tools in the face of a student with less time, less freedom, less confidence and more incentive to use it, what are we giving them to make them want to resist, to commit to their learning, to become something other than a meat puppet plugging syntax into a machine with the machine spewing more syntax out?
At this point, where is the evidence the companies do not wish us harm?
New research suggests that the Department of Government Efficiency has been making inaccurate claims about the extent of its savings from cuts to the Department of Education.
DOGE previously posted on X that it ended 89 contracts from the Education Department’s research arm, the Institute of Education Sciences, worth $881 million. But an analysis released Wednesday by the left-wing think tank New America found that these contracts were worth about $676 million—roughly $200 million less than DOGE claimed. DOGE’s “Wall of Receipts” website, where it tracks its cuts, later suggested the savings from 104 Education Department contracts came out to a more modest $500 million.
New America also asserted that DOGE is losing money, given that the government had already spent almost $400 million on the now-terminated Institute of Education Sciences contracts, meaning those funds have gone to waste.
“Research cannot be undone, and statistics cannot be uncollected. Instead, they will likely sit on a computer somewhere untouched,” New America researchers wrote in a blog post about their findings.
In a separate analysis shared last week, the American Enterprise Institute, a right-leaning think tank, also called into question DOGE’s claims about its Education Department cuts.
Nat Malkus, senior fellow and deputy director of education policy studies at AEI, compared DOGE’s contract values with the department’s listed values and found they “seldom matched” and DOGE’s values were “always higher,” among other problems with DOGE’s data.
“DOGE has an unprecedented opportunity to cut waste and bloat,” Malkus said in a post about his research. “However, the sloppy work shown so far should give pause to even its most sympathetic defenders.”
On Valentine’s Day, the Trump administration surprised schools and colleges with its newest attack on DEI and student body diversity. The U.S. Department of Education’s Office for Civil Rights released a Dear Colleague letter that warned schools and colleges that they may lose federal funding if they discriminate on the basis of race.
This letter revealed novel, unsupported legal theories regarding the application of federal civil rights laws to schools and colleges. In fact, OCR’s letter sweeps so broadly that it claims to prohibit certain considerations of race that remain perfectly legal under well-established legal doctrine.
While the threat of losing federal funding has been a facet of Title VI of the Civil Rights Act since its passage in 1964, the letter specifically takes aim at DEI programming as well as the use of “race as a factor in admissions, financial aid, hiring, training, and other institutional programming.”
Although the letter includes some correct statements of nondiscrimination law, OCR makes assertions that are troubling and unsupported by sound legal reasoning. As part of the team that wrote OCR’s guidance on this very issue in the wake of the Supreme Court’s ruling in Students for Fair Admissions v. Harvard, I am disturbed by how politics is driving policy guidance that will hurt educational institutions and students from kindergarten through college.
In describing the scope of SFFA, OCR’s latest guidance attempts to smuggle in a legal standard that appears nowhere in the court’s opinion. The letter states, “Relying on non-racial information as a proxy for race, and making decisions based on that information, violates the law … It would, for instance, be unlawful for an educational institution to eliminate standardized testing to achieve a desired racial balance or to increase racial diversity.”
Here, OCR baselessly claims that not only can colleges not consider race as a factor in admissions, they also cannot make race-neutral changes to admissions policies that help increase student body diversity—such as eliminating standardized testing. That claim falls firmly outside not only the bounds of SFFA but also the decades of Supreme Court case law that precede it.
In Grutter (2003), Justice Sandra Day O’Connor considers whether the University of Michigan Law School could use a lottery system for admissions. In Fisher (2016), Justice Anthony Kennedy implicitly approves of the Texas top 10 percent plan, perhaps the most well-known race-neutral strategy to increase racial diversity. And in SFFA (2023), the plaintiff’s briefs themselves include endorsements of possible race-neutral alternatives Harvard could have legally pursued such as adopting socioeconomic preferences in admissions.
Yet in its most recent letter, OCR attempts quite the head fake in its declaration that SFFA dictates that schools and colleges must abandon race-neutral strategies meant to increase student body diversity. While in reality SFFA says nothing about the permissibility of these race-neutral strategies, a separate line of cases tackles these legal questions head-on—and contradicts the Trump administration’s unfounded guidance.
In Coalition for TJ, Boston Parent Coalition and other recent cases, groups similar to Students for Fair Admissions have challenged changes to admissions policies of prestigious, selective high schools that were adopted in part to increase student body diversity. In some cases, the schools reconfigured weighting for standardized tests; in others, schools guaranteed that each feeding middle school gets a certain number of seats. In all of the cases, the school districts won. The position now advanced by OCR in its recent letter has failed to find footing in two courts of appeal. And just last year, the Supreme Court declined to further review the decisions in TJ and Boston.
What OCR attempts to do with its letter is extraordinary. It tries to advance a legal theory with support from a Supreme Court case that says nothing about the matter. At the same time, OCR ignores recent judicial opinions in cases that directly address this question.
Regardless of how legally infirm OCR’s proclamations are, schools and colleges will likely feel forced to comply. This could mean that the threat alone will lead schools and colleges to cut efforts to legally pursue racially diverse student bodies and racially inclusive campus environments. As a result, our nation’s classrooms and campuses will unfortunately look less like the communities that they sit in and serve, all because of shoddy policymaking and legal sleight of hand.
Ray Li is a civil rights attorney focusing on education policy. He recently left the Department of Education’s Office for Civil Rights after serving as a career attorney from 2021 to 2025. In that role, he worked on more than a dozen policy documents for OCR, including guidance issued after the Supreme Court’s decision in SFFA. He also served as OCR’s lead staff attorney on appellate and Supreme Court litigation matters, including for the SFFA, Coalition for TJ and Boston Parent Coalition cases. Prior to joining OCR, he advised schools, colleges and universities on legal regulatory issues, including civil rights issues, at Hogan Lovells’ education practice.
Mary Curnock Cook CBE chairs the Emerge/Jisc HE Edtech Advisory Board, and Bess Brennan is Chief of University Partnerships with Cadmus. Cadmus is running a series of collaborative roundtables with UK university leaders about the challenges and opportunities of generative AI in higher education.
Yesterday, Wednesday 26th February, HEPI and Kortext published the Student Generative AI Survey 2025: you can read that here.
Clarity and consistency – that’s what students want. Amid all the noise around Generative AI and assessment integrity, students are hugely concerned about the risk of inadvertent academic misconduct due to misunderstandings within their institutions around student use of GenAI.
That was the message coming loud and clear from the HE leaders at Cadmus’ latest invite-only roundtable, which included contributions from LSE, King’s College London, University of Exeter, Maynooth University and the QAA.
As Eve Alcock, Director of Public Affairs at QAA, put it:
‘Where there’s lack of clarity and uncertainty, student anxiety goes up enormously because they want to do what’s right. They want to engage in their assessments and their learning honestly. And, without clarity, they can’t be sure that they are doing that.‘
Leaders shared examples of good practice around how universities are working in partnership with students to understand how they are using GenAI, as well as their concerns and what faculty and leadership need to know to encourage greater clarity and consistency.
LSE: Student use of AI
At LSE, student use of GenAI has been the focus of a major research project, GENIAL. It’s a cross-departmental initiative to explore the students’ perceptions and experiences of GenAI in their own learning process.
According to Professor Emma McCoy, Vice President and Pro-Vice Chancellor (Education), AI is being used widely by LSE students on the courses covered by the project, and they are specifically using it to enhance their learning.
However, there were risks when AI was brought into modules as part of the project: when it was introduced too early, some students lacked the foundational skills to use it effectively. Those who relied on it heavily in the formative stages generally didn’t do so well in the summative. Using AI tools sped up coding for debugging in data science courses, for example, but there were also examples of AI getting it wrong early on and students taking a wild goose chase because they lacked the foundational skills to recognise the initial errors. In addition, despite banning uploads, students uploaded a variety of copyrighted materials.
While LSE has comprehensive policies around when and how students can use GenAI tools and how it should be acknowledged, only a predicted 40% acknowledged AI use in formative assessments in the project. In any case, such policies may become quickly redundant in any university, warned Professor McCoy:
‘We’re already seeing AI tools being embedded across most platforms so it’s going to be almost impossible for people to distinguish whether they’ve actually used AI themselves or not.‘
Maynooth University: Co-creating for clarity
For Maynooth University in Ireland, the starting point has been the principle that AI is for “everyone but not everything”. Leaders were aware that students were feeling nervous and uncertain about what they could use legitimately, with contradictory guidance sometimes being given about the use of tools with embedded AI functionality, such as Grammarly.
Students were also concerned about a perceived lack of transparency around the use of AI. They felt they were being asked to put great effort into demonstrating that they were not using AI in their work but their openness wasn’t necessarily being reciprocated by the academic staff.
Maynooth’s answer was to set up an expert and genuinely cross-disciplinary AI advisory group to work on complementary student and staff guidelines. Crucially, they brought on board a large group of students to work on the project. From undergraduates through to PhD students and across faculties, students worked in sprint relays on the guidelines, building the “confidence to be able to contribute to this work alongside their academic colleagues to try and remove some of the hierarchies that were sometimes in place,” said Professor Tim Thompson, Vice-President (Students and Learning).
One result was greater clarity around when GenAI use is permitted, with a co-created policy setting out parameters from no GenAI permitted through to mandated GenAI use and referencing.
University of Exeter: Incubating innovation
At the University of Exeter, new guidelines similarly describe every assessment as either AI integrated, where GenAI is a significant part of the assessment, AI supported, where ethical use of GenAI is accepted along with appropriate transparency, or AI prohibited. Alongside clear calls for students to be open about their GenAI use is a push for staff to also show best practice and be transparent with students if they are using GenAI to monitor or assess their work.
To that end, Exeter is supporting staff to pick up the challenge of integrating GenAI into learning and teaching to better understand the ways in which students are accessing these tools. Exeter’s Education Incubator offers opportunities for staff to explore pedagogic innovation in partnership with students, with the intention of scaling up successful interventions. Projects include rethinking historical skills assessment with LLMs in Classics and Ancient History and student-led hackathons on detecting bias in AIs.
While recognising the opportunities offered by AI for democratising access to learning, with innovations such as 24/7 coaching, Professor Tim Quine, Vice-President and Deputy Vice-Chancellor Education and Student Experience, also highlighted the danger of ever-widening digital divides:
‘There is a significant risk that AI will open up gaps in access to support where privileged students can access premium products that are inaccessible to others, while universities are struggling to navigate the legal, ethical and IP issues associated with institutional AI-supportive technologies, even if we’ve got the budgets to put those technologies in place.‘
King’s College London: rethinking assessment
‘No one wants more assessment!‘ declared Professor Samantha Smidt, Academic Director, King’s College London.
‘Talking to staff and to students about assessment, they’re asking for things to be different, to be more meaningful and rewarding, but nobody’s saying we don’t have enough of it.’
King’s has grasped the opportunity to rethink assessment in a holistic way as a result of, but not limited to, the AI imperative. Aware that students are keen for credit-focused work to be transparently comparable across modules, and for marks to be comparable across markers, King’s has been working on culture change, socialising ideas around a new approach to assessment.
Professional development opportunities – and, importantly, funding, for staff to explore areas of interest within a new assessment framework, TASK – have been allied with student partnership work. Pilots are trialling programmatic assessment across 10 programmes, focusing on assessment timings, tariffs, a shift to more formative assessment and a drive to reduce turnaround times.
In the loop
AI policy in HE is an increasingly complex area, with staff AI literacy requiring as much attention as student use of GenAI tools. In such a fast-developing field, where AI is going to become ever more woven into the fabric of the everyday technology used in academic work, engaging staff and students with emerging issues is critical.
‘The thing we hear time and time again is to keep partnering with and learning from students as a continuous process,‘ urged QAA’s Eve Alcock:
Some AI policies will have been in place for a year, two years now, which is brilliant, but are they still working? Is there a need to evolve them? Making sure that students are fully within that loop is incredibly important.
