Core Learning Exchange

A new randomized trial found that a course AI tutor lowered college students' final grades. The reason is worth every school's attention.

At the University of Maryland, the AI tutor came set to give direct answers, with a guided tutoring mode available. Only two of 30 instructors ended the term using guided mode. Students used it mostly to get answers, participated less in class activities, and first-generation students lost the most ground. The lesson: the default setting is the teaching method.

Read our analysis here:
https://www.corelearningexchange.com/blog/ai-tutor-default-settings-2026.html

An AI Tutor Lowered Grades. Blame the Default Setting A semester-long Maryland trial found a course AI tutor cut final grades 0.37 SD. Most instructors left it in answer mode. Configuration decides outcomes.

"College or apprenticeship" is the wrong question, and new data proves it.

A Federal Reserve Bank of Philadelphia study found that about half of apprentices attended college before starting, and about 40 percent of those who started with no college went on to earn a bachelor's degree or higher. Learners move back and forth between these paths all the time. The problem is that their certifications and coursework often do not count on the other side. We looked at what CTE programs can do to build pathways that travel with students.

Read the full analysis:
https://www.corelearningexchange.com/blog/apprenticeship-college-pathways-2026.html

Apprenticeship vs. College Is a False Choice Apprentices often move through college before, during, and after their programs. A new Philadelphia Fed study shows why CTE must build pathways, not silos.

AI tutors now match expert human tutors at a fraction of the cost. So why is that not the whole story?

A new study of 2,383 learners found AI tutoring produced the same one-hour GRE gains as expert humans at 918 times lower cost. But those learners were motivated volunteers. In real classrooms, research shows up to 47 percent of students assigned an AI tutor never use it. We break down the five questions every school should ask before trusting a cost-per-gain claim.

Read the analysis here:
https://www.corelearningexchange.com/blog/ai-tutoring-cost-per-gain-2026.html

AI Tutoring Got 918x Cheaper. Measure What Matters A new study finds AI tutors match expert humans on GRE gains at 918x lower cost. The harder question for schools is who keeps showing up after hour one.

Are CTE students failing the credential, or failing the calendar?

Pittsburgh Public Schools runs more than 60 CTE programs, yet only about 19 percent of seniors earned an industry credential last year. The district's plan to raise that number starts with a simple change: more than one chance to take the exam. When a certification test comes once at the end of senior year, a near miss means graduating with nothing to show an employer. We looked at what makes a second attempt actually count.

Read the analysis:
https://www.corelearningexchange.com/blog/cte-credential-retake-windows-2026.html

One Test, One Shot Is Why CTE Credential Rates Stall Only 19% of Pittsburgh seniors earn an industry credential across 60+ CTE programs. Its fix, a second test window, shows why single-shot exams stall CTE.

Two-thirds of Harvard faculty now say AI has had a negative effect on their courses. Last year it was 42 percent.

That is not a training problem. Those professors watched these tools work inside their own classrooms for two years and made a call. What they are rejecting is AI that shows up from outside the course, knows nothing about the syllabus, and answers to no one. In a course, the struggle is the instruction. A tool that removes it removes the teaching.

The fix is not better detection. It is making student thinking visible again.

Read the full analysis:
https://www.corelearningexchange.com/blog/faculty-ai-backlash-2026.html

Faculty Are Not Anti-AI. They Are Anti-Unbounded AI. Two-thirds of Harvard faculty now say AI hurt their courses, up from 42% a year ago. The verdict is not on AI. It is on tools that sit outside the course.

A CTE instructor spends about $500 of his own money every year to keep the industry credentials he teaches toward. His district does not reimburse a dollar of it.

New Education Week reporting puts numbers on a gap the field has absorbed for years. Districts fund the recruiting of industry professionals into classrooms. Almost none fund keeping those professionals current once they arrive. The teacher shortage is visible. Credential decay is not, and it is the more expensive problem.

Read what districts can actually do about it now.

https://www.corelearningexchange.com/blog/cte-teacher-industry-currency-2026.html

CTE Teachers Are Funding Their Own Relevance CTE enrollment is surging while instructors self-fund recertification. Districts budget to recruit industry pros but not to keep their skills current.

New York City just paused student-facing AI for 600,000 students through eighth grade, and the policy is more interesting than the headline.

Teachers keep AI for lesson planning. High schoolers keep it for AI literacy and career readiness. What the district cut was software that puts a general-purpose chatbot in front of a twelve-year-old with no instructional boundary, no teacher visibility, and no record of what the student actually learned. That is a product critique, not a verdict on AI in schools.

Read the full analysis of what districts should be asking vendors instead.

https://www.corelearningexchange.com/blog/district-ai-bans-bounded-tutoring-2026.html

NYC's AI Ban Is a Product Critique New York City paused student-facing AI for 600,000 students through grade 8. The policy is not anti-AI. It is a verdict on unbounded chatbots.

The CTE research that made expansion politically easy came from dedicated CTE schools, not from added pathways at comprehensive high schools.

New Massachusetts research covered by Chalkbeat found that when comprehensive high schools expanded CTE, graduation rates showed little improvement, four-year college enrollment fell, and the early earnings advantage faded by age 27. The difference is not the course title. It is the employer partnerships, the instructional depth, and whether anyone can measure program quality before the first cohort graduates.

Read the full analysis of what districts should measure in year one.

https://www.corelearningexchange.com/blog/cte-expansion-replication-gap-2026.html

Expanding CTE Is Not the Same as Replicating It New Massachusetts research finds expanded CTE at comprehensive high schools did not lift graduation rates, and early earnings gains faded by age 27.

Two in three highly AI-exposed occupations are traps.

A Bipartisan Policy Center brief out today finds that when workers leave a highly AI-exposed job, the jobs they most commonly move into are just as exposed. That is roughly 16.7 million people, about 10 percent of employed Americans. Retraining that moves someone sideways is not a pathway. The brief argues the real problem is not the disruption itself, it is that our education and workforce systems were designed to adapt on a timeline the economy no longer honors.

Read the full analysis on the Core-LX blog.

https://www.corelearningexchange.com/blog/workforce-system-redesign-2026.html

16.7 Million Workers Have Nowhere to Go A Bipartisan Policy Center brief finds two in three AI-exposed jobs are trapped. Retraining that moves workers sideways is not a pathway. Here is the fix.

The largest school district in the country just paused its AI rules, but the deadlines facing everyone else are not pausing.

New York City delayed its school AI guidance after 6,500 public comments, while 77 AI-in-education bills move through 27 states. Buried in the retreat was the clearest standard a district could ask for: prepare students for a world where AI is present without letting it replace their thinking. That is a testable line. A tool either does a student's thinking or it keeps the student doing it.

Read why the policy vacuum is not a reason to wait.

https://www.corelearningexchange.com/blog/ai-classroom-policy-vacuum-2026

Core Learning Exchange - Equitable pathways to career credentials Comprehensive CTE solutions for grades 6-14. Over 450 courses from 20+ providers, aligned to state standards and industry certifications.