Filtered by tag: academic integrity × Clear
AI Task Forces Drive Policy, Adoption and Innovation in Higher Ed
Matched: Artificial Intelligence
Colleges are forming cross-functional AI task forces to address challenges outpacing governance, including shadow AI use by departments operating outside IT oversight, equity gaps between students with access to premium versus free AI tools, and faculty struggles to uphold academic integrity without becoming full-time AI-detection enforcers.
AI and the humanities: crisis or opportunity?
Matched: Artificial Intelligence, generative AI
AI poses both a threat and opportunity to humanities education. LLMs risk devaluing degrees if students submit AI-generated work — professors identified it correctly only 53% of the time in one study. Yet graduates who develop genuinely human skills like critical thinking, ethical reasoning, and cultural understanding may become more valuable to employers precisely because AI cannot replicate these abilities. Universities should adopt assessments, such as oral exams, that cannot be AI-generated.
WEEKEND READING: When academic misconduct becomes an identity: what happens after the sanction?
Matched: generative AI
Academic misconduct procedures focus on detection, investigation, and sanction but rarely address what comes next for students. Dr Manesha Peiris argues institutions should consider reintegration — allowing students to demonstrate learning and rebuild trust without past findings permanently shaping how their work is viewed. She also raises equity concerns around GenAI, warning that markers of "suspicious" work may disadvantage some students disproportionately.
Would You Hire a Ph.D. Who Wrote Their Dissertation With AI?
Matched: AI
A new working paper found AI writing tools appearing in nearly one-third of STEM doctoral dissertations filed between January and May of this year, raising questions about whether employers and academic institutions should be concerned about hiring PhD graduates who used AI to help write their dissertations.
How to Assign Research Papers Again
Matched: AI
Two political science professors argue that research paper assignments remain valuable but require adaptation in the AI era. They suggest three approaches: requiring students to submit drafts and process documentation, conducting brief oral follow-ups where students discuss their written work, and designing more specific, personalized prompts that are harder for AI to address generically. The goal is restoring confidence in authentic student writing.
Is it time to ditch AI detection tools?
Matched: AI, Artificial Intelligence
AI detection tools used in education are unreliable and should be abandoned, argues a Charles Sturt University researcher. Even low false-positive rates can wrongly flag thousands of students for misconduct, causing significant stress. Tools also struggle with real-world contexts where AI assistance blurs into human writing. Instead, educators should redesign assessments to include supervised tasks that genuinely measure student learning outcomes.
As more Australian universities use AI to mark students’ work, is it creating a ‘slop-cycle’?
Matched: AI, Artificial Intelligence
Australian universities are increasingly using AI to help mark student assignments, raising concerns about a damaging feedback loop. Academics warn that if AI grades work, students may use AI to write it, producing low-quality "slop" that AI then evaluates — a cycle that could undermine genuine learning. Critics say the trend poses an existential threat to higher education, though institutions are divided on how to respond.
AI Writing Detection: A Losing Battle Worth Fighting
Matched: AI
Tools designed to detect AI-generated text are proving unreliable, frequently flagging human writing as machine-made and missing actual AI content. Experts disagree on whether the technology can ever be accurate enough to be useful. Despite flaws, many educators and institutions continue using detectors, arguing that imperfect tools still create some deterrent, even as critics warn the false positives cause real harm to students.
Dartmouth’s Provost Quietly Used AI in His Published Work. It’s Touched Off a Campus Firestorm.
Matched: AI
Dartmouth's provost acknowledged using AI-generated text in a published academic article without disclosure, sparking controversy on campus. Critics argue the incident undermines academic integrity standards students are held to. The situation has intensified broader debates at Dartmouth over the college's generally permissive stance toward AI use, with faculty and students questioning whether leadership is modeling appropriate behavior.
Universities criticised for OpenAI deals
Matched: AI
Two University of Sydney and UNSW have partnered with OpenAI to integrate ChatGPT Edu into learning and research, drawing criticism from the National Tertiary Education Union. NSW division secretary Vince Caughley condemned the deals as "shameful and reckless," citing poor staff consultation, weak governance, and risks of locking institutions into a contract with a loss-making vendor. Both universities defended the partnerships as preparing students for an AI-shaped workforce.
7 Practical Ways to Use AI as a Thinking Partner in Your Classroom
Matched: AI
Philosophy professor Dr. Jessa Roisen advocates using AI as a thinking partner rather than banning it. Her seven strategies include clarifying assignment goals, having students critique AI-generated responses, using AI to test arguments, building in reflection, and creating low-stakes practice. She emphasizes anchoring AI to course materials and notes that while AI can generate text, it cannot replace the critical thinking students develop by genuinely wrestling with ideas.
College students are writing with AI – but a pilot study finds they’re not simply letting it write for them
Matched: AI, Artificial Intelligence
A pilot study at Kennesaw State University used think-aloud protocols and screen recordings to observe how 20 undergraduates actually used AI while writing, rather than relying on surveys or finished papers. Researchers left the room to reduce self-conscious behavior. Students typically used AI for brainstorming and overcoming writer's block, but frequently edited or rejected its output, maintaining control over argument and voice. The findings suggest AI functions more as a collaborative tool than a ghostwriter, though researchers caution the small sample limits broader conclusions.
Generative AI in business schools: friend or foe?
Matched: AI, Artificial Intelligence
A UK study of 15 business graduates found students use ChatGPT routinely for drafting, summarising, and overcoming writer's block, but harbour nuanced concerns. Key tensions include unequal access to premium AI tools versus AI's potential to support students with disabilities or language barriers, and inconsistent institutional rules creating anxiety around academic integrity. Students also fear employers may undervalue their degrees as "AI-generated." Researchers recommend clearer guidelines, process-focused assessments, and equitable access to AI tools.
A writing professor’s new task in the age of AI: Teaching students when to struggle
Matched: AI
A writing professor describes navigating AI in higher education, neither banning it nor embracing it uncritically. Early classroom experiments revealed AI's errors and biases. Over time, she shifted from neutral observer to guide, asking students to write without AI so they can understand their own thinking first. Research shows AI use can improve short-term scores while hindering deeper learning. Her approach focuses on building student judgment about when AI helps versus when it undermines genuine intellectual development.
The greatest risk of AI in higher education isn’t cheating – it’s the erosion of learning itself
Matched: AI, Artificial Intelligence
The real danger of AI in higher education isn't cheating but the erosion of learning itself, argue researchers from UMass Boston. As AI handles more teaching, research, and administrative tasks, universities risk hollowing out the mentorship pipelines and productive struggles that build genuine expertise. From automated advising to agentic research tools, the shift raises questions about transparency, accountability, and whether universities remain ecosystems that form human judgment or merely credential factories.
Student cheating dominates talk of generative AI in higher ed, but universities and tech companies face ethical issues too
Matched: AI, Artificial Intelligence
Debates about AI in higher education focus mainly on student cheating, but broader ethical issues deserve attention. These include tech companies training AI models on copyrighted material, student data privacy under vendor agreements, and unequal access to AI tools. Experts argue responsibility shouldn't fall solely on students — universities should vet AI tools, disclose vendor terms transparently, and update academic integrity policies, while tech companies must address their own ethical violations.
Where does human thinking end and AI begin? An AI authorship protocol aims to show the difference
Matched: AI, Artificial Intelligence
A philosophy professor argues that AI threatens accountability in education and professions by obscuring whether humans actually did the thinking behind submitted work. He is piloting an "authorship protocol" at Temple University that uses conversational AI prompts to verify student reasoning aligns with their writing, without keystroke surveillance. The system issues secure tags certifying work met specified conditions, aiming to preserve human judgment rather than detect AI use.
At one elite college, over 80% of students now use AI – but it’s not all about outsourcing their work
Matched: AI, Artificial Intelligence
A Middlebury College survey of 634 students found over 80% use generative AI for coursework, with most using it to enhance learning rather than cheat. Students most commonly used AI to explain concepts, describing it as an on-demand tutor. While 42% used AI for automation tasks like essay writing, this typically occurred during high-pressure periods or for low-stakes work. Findings were consistent with global data and verified against Anthropic's actual usage logs.