California Community Colleges showcase AI pilots to detect enrollment fraud, speed apprenticeship approvals and help students

Presenters at the Board of Governors meeting described multiple artificial intelligence projects now operating at California community colleges that system leaders say have reduced fraudulent enrollments, shortened backlogs in apprenticeship degree audits and are being piloted to provide faster student supports.

Jory Hadsell, executive in residence at the chancellor’s office, described how an AI fraud‑detection model co‑developed with End to End Services (branded in presentations as Light Leap AI) began as a local solution at Foothill‑De Anza and spread by demonstration to West Valley‑Mission and Santiago Canyon. "From those 3 colleges, adoption spread quietly and quickly…as of today, 48 colleges in our system have locally adopted this solution," Hadsell said.

Santiago Canyon College President Jeannie Kim said her college removed roughly 10,000 fraudulent enrollments last fall, and after implementing the vendor model and verification steps the college enrolled about 8,000 legitimate students in those seats. "We have a 99 percent efficacy rate with the implementation that we have done," Kim said, describing a multilevel verification approach that includes ID checks and bilateral information sharing across colleges.

Kiran Koditala, chief executive of End to End Services, told the board the company and partners have processed "close to 3,000,000 admissions applications, and we caught over 360,000 fraudsters," numbers he said were produced in the product’s recent deployments. District staff credited the chancellor’s office Digital Center for Innovation and work on common integrations for enabling faster, secure rollouts across districts with different student information systems.

Speakers also highlighted other AI‑powered tools: Santiago Canyon described a newly developed apprenticeship‑tracking assistant that will reduce a three‑year manual backlog for 7,500 apprentices; presenters said the system has already processed several hundred degree and certificate substitution cases in beta. End to End and partners showed how OCR and large‑language‑model techniques are being tested to ingest handwritten applications from incarcerated learners and to map courses across colleges for the California Virtual Campus.

Board members praised the work but asked detailed questions about governance, student data protection and whether smaller, rural colleges will be able to adopt the tools. Faculty and the Faculty Association of California Community Colleges urged stronger shared governance and consultation: "Consultation council and other stakeholder groups can offer valuable perspectives on how to structure industry partnerships, safeguard student data, and ensure new tools truly serve the educational experience," said Wendy Burlweinkoop, president of FACC, during public comment.

A virtual public commenter questioned vendor privacy policies and urged local control over data: "No student should be forced to give highly personal information to another corporation," said Omar Zavala, who asked for full disclosure of verification partners and retention policies.

Chancellor’s office staff said data governance and the statewide AI council’s "human‑in‑the‑loop" principles guide pilots and that they will pursue standard privacy, opt‑out and integration policies through a newly expanded data governance advisory work group. Presenters said the goal is to keep humans in critical decisions while using AI to triage high volumes of work and reduce staff burnout.

Provenance: Demonstrations and Q&A occurred during Item 5.6, the AI educational series presentation.