Faculty development in the AI era: what an effective FDP looks like

Generative AI arrived in Indian classrooms through students first. Assignments began to look polished in new ways, and faculty were left to decide on their own whether to ban, ignore or adopt tools like ChatGPT and Gemini. Many institutions responded with a quick FDP: a guest lecture, a demonstration of prompts, a certificate, and a line in the AQAR.

That format is easy to organise and rarely changes practice. Faculty return to the same syllabus, the same assessment pattern and the same uncertainty. An effective faculty development programme for the AI era looks different: it is longer, hands-on, discipline-specific and judged by what changes in courses afterwards.

What faculty actually need

Before designing an FDP, ask faculty what they find hard. Across disciplines, the needs tend to fall into four groups.

• Confidence with the tools: what current AI assistants can and cannot do reliably, and how to use them safely for preparation and feedback.
• Assessment redesign: how to set assignments that still measure learning when students have AI access.
• Pedagogy: how to use AI to support active learning, practice and feedback without replacing thinking.
• Policy and ethics: what the institution allows, how to handle suspected misuse fairly, and how to protect student data.

Generic sessions on 'AI in education' cover the first group lightly and skip the rest. The second and third groups are where teaching changes.

Design principles for an FDP that works

1. Define outcomes for faculty. Write what participants will be able to do at the end, in the same way you would write course outcomes. For example: redesign one assessment in their course to be AI-resilient, or build a structured AI-assisted practice activity.
2. Spread it out. A series of shorter sessions over several weeks, with work between sessions, beats a single intensive day.
3. Make it discipline-specific. A chemistry lab, a law moot and a marketing case use AI differently. Group participants by discipline for practical work.
4. Work on real courses. Participants should bring a current syllabus and leave with revised materials for it.
5. Include peer review. Faculty review each other's redesigned assessments and activities against agreed criteria.
6. Follow up in the semester. A check-in after the redesigned course has run captures what worked and what did not.

A sample structure

The structure below is a template that institutions can adapt. Durations are indicative.

Assessment redesign deserves the most time

If an assignment can be completed entirely by pasting the question into an AI tool, it no longer measures what it was designed to measure. Faculty need practical alternatives, and the FDP should give them time to build them.

• Ask for process evidence: drafts, notes, version history or reflection on how the work developed.
• Use local, recent or personal data that general tools cannot answer from memory, such as a field survey or the student's own lab readings.
• Add a short oral component in which students explain and defend their work.
• Move some assessment into supervised class time, especially for foundational skills.
• Where AI use is permitted, require students to disclose how they used it and assess the quality of their judgement in using it.

Measuring whether the FDP worked

Attendance and feedback forms show that an FDP happened. They do not show that it worked. Track measures that relate to practice and learning.

• Number of courses with revised assessments and AI use statements in the next semester.
• Quality of revised materials, reviewed by department heads or IQAC against a rubric.
• Student feedback on clarity of AI rules and usefulness of AI-supported activities.
• Changes in CO attainment for courses where assessment was redesigned, interpreted with care.
• Faculty confidence before and after, measured with the same short survey.

Department heads play a large part here. When the FDP ends, someone has to ask in the next department meeting which courses have changed and what students made of it. Building that review into the department calendar, rather than leaving it to individual enthusiasm, is often the difference between a programme that changes teaching and one that is forgotten within a semester.

These measures also produce useful evidence for accreditation. NAAC's governance and teaching-learning criteria consider faculty development and its effect, and records of revised courses tell a stronger story than a list of certificates.

Common pitfalls

A few mistakes recur. Tool-only training dates quickly as products change. Sessions run by external speakers with no follow-up leave faculty without support when they try ideas in class. Mandatory attendance without protected time breeds resentment. And programmes that treat AI only as a threat to integrity miss its potential to give students more practice and faster feedback. Balance matters: faculty should leave knowing both the risks and the specific ways AI can help their students learn.

Another pitfall is ignoring differences in starting points. Some faculty already use AI tools daily; others have never opened one. Mixed groups work for discussion of pedagogy and policy, but hands-on sessions go better when participants are grouped by comfort level, with experienced colleagues acting as peer mentors. Senior faculty in particular often respond better to a peer demonstrating a working classroom example than to an external presentation.

How MSM CampusOS can help

MSM CampusOS designs and runs discipline-specific faculty development with follow-up through the semester. Our FDP on AI and digital teaching is built around redesigned courses and assessments, and links to our wider faculty development offering. To plan a programme for your departments, book a demo.