Forbes India

Has the curricula prepared this generation for the job market AI has rewritten?

Mohit Anand was nine days into his first semester at Netaji Subhas University of Technology, Delhi, in November 2022 when OpenAI launched ChatGPT, forever changing the way work would be done and, therefore, the skills employers would expect from new hires.The final-year engineering student at the Delhi state university did not see a lot of AI integrated into his college curriculum. But students, he says, were “on top of it”. They found the tools themselves and incorporated them into everyday lives.While most of them used GenAI tools to complete assignments, Anand built an AI startup, which was incubated by his university. He eventually decided he needed corporate experience and went for a job, a task that was easy for him given his credentials of having founded an AI venture.In neighbouring Haryana, Pratyaksh Saluja, who is a year junior to Anand, walked into his internship interview this year with something no engineering student had ever needed to bring before: A screen recording.Earlier engineering students would show their GitHub profile at interviews, which lets recruiters assess a student’s coding skills and see how well they are contributing to an actual project.But the BML Munjal University student says these days there is a demand for innovative projects. “At interviews, I show them recordings of my projects which include AI agents, workflows, channels like Slack etc. So, the showcase of work is a mix of both now,” says the third-year student at the private university.Anand and Saluja represent India’s in-between generation of engineers—students who enrolled before GenAI was mainstream and are graduating into a world that can’t operate without it. They are doing all of this at possibly one of the worst times to be looking for a job in information technology (IT).IT—the sector that for two decades served as the great absorber of engineering talent—is facing an uncertain demand environment and big companies are pulling back, or rethinking their fresher hiring targets.IT major TCS has rolled out offers to about 25,000 for FY27 compared to 44,000 in the previous year and plans on revising targets depending on the demand. Infosys will hire 20,000 freshers, on par with last year while TechM, HCLTech and Wipro have adopted a wait-and-watch strategy for the fiscal.Engineering students are stepping into this market carrying credentials from a curriculum that—everyone from recruiters, industry leaders and educators agree—has not kept pace with what the industry now wants.Engineering curricula need an overhaul, according to stakeholders.Kamlesh Vyas, a partner at Deloitte India, says the core of what’s taught in engineering colleges hasn’t changed but the skill application has shifted dramatically. “What employers are looking for is an AI-plus-X kind of approach. Whatever you do, you’ve got to do that well plus you also think about applying AI to that.”Most curricula, he says, don’t get changed drastically and by the time they do, it’s too late.The story is better at top-tier colleges such as the Indian Institutes of Technology (IITs), but only slightly.Raunit Patel was better prepared than most. The final-year student in BTech data science and AI at IIT-Guwahati’s Mehta Family School of Data Science and AI—one of the few institutions in India to have established a dedicated AI school, in 2021—is part of a cohort whose placement outcomes had some of the best names, from Google and Microsoft to JPMorgan Chase, Oracle and BNY Mellon.But he is under no illusion that the curriculum alone is enough. “To keep up with ongoing advancements such as agentic AI, RAG systems and multimodal LLMs, students often take additional certifications, follow online resources and use platforms like YouTube for practical understanding,” says Patel.Students from top-tier institutions, especially those where AI is being integrated meaningfully into curriculum, pedagogy and assessment, are being absorbed into GCCs, product companies, startups, analytics roles and AI-adjacent functions commanding premium compensation, says Avantika Tomar, partner in education sector practice at EY-Parthenon. “But the middle- and lower-tier colleges are struggling because employers are increasingly asking for evidence of readiness, not just credentials or degrees.”That is the truth this generation is facing: The institutions that produced them—with some exceptions—were not built for the speed at which this technology arrived. And the onus is on the students to get AI-literate on their own time.Let’s look at some numbers.Kapil Joshi, CEO of IT staffing at Quess Corp, says India today has a core AI workforce of only 350,000 to 380,000 professionals against an immediate demand for 680,000 to 710,000 roles. “And the gap is most visible in job-ready talent rather than awareness alone,” he says. “When only about 15 percent of engineering faculty have production AI exposure, it is clear that industry and academia can no longer function in silos.”He calls GenAI a permanent reset in how employability and engineering talent will be defined.EY-Parthenon’s Tomar is equally direct. “A curriculum designed for a services economy of early 2000s cannot remain adequate in an AI-driven, platform-centric, interdisciplinary market.”Ankit Aggarwal, founder and CEO of hiring platform Unstop, has been watching this gap widen in real time. “Both recruiters and college directors agree that the college curriculum is at least five years behind industry trends,” he says.Aggarwal highlights a data point that shows how colleges lag in preparing students. Amazon started the trend of hiring engineers for product companies; before that, hiring was services-focussed. “But how many colleges are teaching product management as a subject today? Less than 1 percent. Same is the case with AI literacy.”When his platform tested students specifically on AI literacy—not usage—the results were sobering. “Not even 20 percent cleared the test,” he says. “More than 50 percent of the students say they didn’t formally receive any education on AI literacy from their college. And that is true for all, including the IITs.”The roles being hired for today barely resemble what they were three years ago. Viswanath PS, managing director and CEO of Randstad India, says three years ago, an entry-level engineer was hired to write modular code or manage basic data sets. “Today, the requirement is for systemic thinking. Employers now demand that freshers understand the entire lifecycle of a project, utilising AI to handle the drudge work while the human focuses on architectural integrity, security and ethics.”EY-Parthenon’s Tomar says graduates who can demonstrate project experience, AI fluency, communication ability and applied problem-solving are performing materially better than those with only a conventional academic profile.“Many colleges continue to teach in a theory-heavy, syllabus-led format, while employers are increasingly hiring for applied capability,” says Tomar, who’s also learning and development head at EY-Parthenon. “Many institutions are still teaching syntax, while employers are screening for semantics.”Knowing C, Python or any coding language, or knowing financial concepts, 4Ps of marketing or Porter’s five forces, is no longer the differentiator. “Using those capabilities to solve problems, integrate ideas and produce meaningful outputs is the new baseline,” she adds.Yash Sinha, assistant professor of computer science at BITS Pilani, says the entire hierarchy of what gets taught first needs to be inverted. “Previously, we taught syntax first and architecture last. Now, we must teach architecture first. If an AI can write the function, the student’s primary value lies in determining where that function belongs in a distributed system—and whether it should exist at all.”At BITS, he says, this rethinking is already underway. Students are being trained in what he calls “epistemic checking”, which involves training students to rigorously verify the truth, structure and intent of an AI’s output rather than blindly trusting syntactically correct code. He cites Andrej Karpathy’s observation that vibe coding works, until it doesn’t. “Our responsibility as educators is to teach students what to do when the vibe fails.”Also Read: Why MBA education in the age of AI must move beyond analysis for leadershipEngineers are facing a paradox. “The easier it is to generate code, the harder it is to debug it. When a human writes code, they build a mental map of the logic. When they vibe code, they lack that map,” says Sinha.BITS Pilani is addressing this by shifting their labs from implementation to forensics. “We give students broken, AI-generated codebases and ask them to diagnose the failure. This mimics the real world, where they will likely spend more time reviewing AI code than writing their own,” he adds.If syntax is losing value, what does that mean for the broader category of memory-based knowledge, which is what the bulk of our educational institutes test students on?Tomar says it’s quickly losing relevance as a differentiator. “In a world where graduates have access to AI tools, and increasingly agentic systems supporting their work, rote recall is no longer a competitive advantage. Colleges that continue to over-index on memory-based assessment are likely to fall behind.” She adds: “Sadly, or unfortunately, the bulk of our educational institutes are still testing students on memory and retention.”The shift, she says, is from retention to reasoning. The premium is moving toward “understanding systems, product logic, architecture, debugging, validation and deployment”.This is the gap the in-between generation has had to bridge on its own.Slow ChangeInstitutions, on their part, have been incorporating changes.Snehanshu Saha, head of the Anuradha and Prashanth Palakurthi Centre for Artificial Intelligence Research at BITS Pilani’s Goa campus, is careful to push back against the narrative of institutional inertia. “AI adoption at BITS Pilani is not a buzzword and it’s certainly not a catch-up game,” he says.What that means in practice is engaging alumni from OpenAI, Meta and Google to teach generative AI; launching a minor in computational economics in 2022; building new AI and autonomous systems programmes.“We’ve also started revisiting courses that have traditionally been central to physics and economics majors but have quietly made a significant impact on AI,” says Saha, who is also a professor in the department of computer science & information systems.The process, he acknowledges, may appear slow from the outside. But “it must be meaningful. We’d rather build something that lasts than chase a trend that shifts every six months.”Bhaskar Datta, dean (academics), IIT-Gandhinagar, echoes a version of this. Placement preparation still rests strongly on basics, he says. “What has changed is that students now also need to understand how AI fits into that workflow. We encourage students to think of AI as an augmenting tool, not as a replacement for their own reasoning.”Coding remains relevant. “In fact, I would argue that coding is still one of the clearest ways to learn how to think computationally,” says Datta.Also Read: AI & jobs: Are companies prepared for the workforce revolution?He notes that students returning from placement rounds consistently report that fundamentals still matter far more than they expected. “Many students realise that interviews continue to test core understanding more than superficial familiarity with trends. At the same time, students are also becoming aware that the workplace now expects comfort with modern tools, including AI-assisted workflows.”Even as the institute has started integrating AI in its curriculum—from adding courses to changing existing ones—it encourages students to go beyond the formal curriculum and supplement classroom learning with technical blogs, documentation, open-source projects and practical experimentation.But even granting all of that, the pace problem remains.EY-Parthenon’s Tomar identifies three broad archetypes of institutional response she is seeing across Indian colleges. The first: Non-adopters, “institutions that are still relying on traditional teaching models, syllabus coverage and leaving AI usage almost entirely to students”.The second: Cosmetic adopters, those that have added one or two AI courses but are “still treating AI as a vertical add-on rather than a structural shift”.The third (and the rarest): Real transformers, “institutions that are embedding AI horizontally across curricula, pedagogy, and assessment, and are redesigning the learning model accordingly”.“The third archetype will define the winners,” she says.Onus on StudentsIn the absence of a system-wide overhaul, the in-between generation has built its own parallel education.Nipun Sharma, CEO, TeamLease Degree Apprenticeship, says there is some degree of uncertainty among students, which is expected during a period of rapid technological change.But that uncertainty is translating into something more productive than paralysis. Students are “engaging more actively with learning opportunities, internships and skill-building initiatives”.Randstad’s Viswanath sees this bifurcation playing out directly in the hiring market. “Those who have treated AI as a co-pilot are seeing a 20 to 25 percent premium in interest, while those who simply relied on it as a fall-back face a skills-hollow risk where their foundational logic is perceived as weaker than previous years.”Back at BML University, third-year student Saluja says traditional study is currently one to two hours and the rest is playing around with AI tools. He has shifted to Claude Code almost entirely for his development work, but “pricing has gone up and token maxing is a real thing. So, you have to manage your skills”. He has found a workaround: The startup he’s been building workflows for pays for his tokens.His batchmate Shrey Jaiswal is more philosophical about the upheaval. “At the end of the day, AI can only augment what an engineer does,” he says. “AI is a great solution bridging the gap between thinking and delivering. It cannot think of solutions given real-life constraints. I am not that worried. Every major technology change always leads to restructuring of roles.”Things he’s learning at college will not go to waste, he feels. “It can be applied to n number of iterations after that.”Agrima Agarwal, another of Saluja’s batchmate, says right now recruiters do not expect engineers to write code using AI. “They test how I can scale things, make it better.”Beyond the CampusIncreasingly, by the time a formal interview happens, the most competitive candidates have already differentiated themselves somewhere else entirely.Unstop’s Aggarwal reports hiring intent climbing from 69 percent of companies in 2025 to 88 percent in 2026. “The real shift is in how companies are hiring,” he says. “The emphasis is increasingly on job-ready capabilities, adaptability and real-world problem-solving—prioritising capability over credentials.”That shift has found a natural home in hackathons, conferences and professional networking. Agarwal has been connecting with startup founders directly on LinkedIn. Jaiswal has found that “conferences and events help in finding new opportunities in addition to traditional channels”.Asked if large-scale hackathons are shifting from traditional coding to AI prompt engineering-based competitions, Sinha of BITS Pilani, says, “They already are, but framing this shift as ‘prompt engineering’ undersells what’s really happening. The real change is toward complexity management.”Tomar says internships have moved from optional to essential.The system is moving. But is it moving fast enough?Randstad’s Viswanath says while academic cycles move in four-year blocks, the GenAI landscape moves in four-week sprints.Deloitte’s Vyas says a good professor should be reassessing their teachings daily. “I am, of course, talking at a philosophical level but that’s the speed of change with AI. Teaching has to be dynamic as well.”

Google Play App Store