AI reshapes India's graduate tech hiring and career paths | Magica
AI reshapes India's graduate tech hiring and the route to senior jobs
Editorial Team
••📖7 min read
India's IT lobby chief warns that AI could make technology jobs harder for graduates to secure. Hiring is below earlier peaks, but some firms favour AI-equipped beginners over costly experienced staff. The unresolved question is how junior workers will build expertise.
India's IT lobby chief warns that automating coding and testing could make graduate job searches harder.
Graduate intake has fallen from earlier peaks, even as some firms favour AI-equipped beginners over expensive experienced hires.
Research identifies a risk to junior workers' training—but also evidence that AI assistance can accelerate learning.
Recent college graduates in India could struggle to find technology jobs as companies automate coding and software testing, Rajesh Nambiar, president of technology-industry body Nasscom and former India chairman and managing director of technology-services company Cognizant, warned in a November 6 interview.
The warning raises two linked decisions for employers: how many graduates to recruit, and how to develop their judgment when AI performs more of the routine work that once taught them the job. The evidence points to changing entry routes rather than a uniform retreat from hiring beginners.
Estimated technology employment in 2026 and the company-reported Infosys graduate recruitment decline cover different populations and periods. Neither measures jobs lost to AI. Source: CNA.
Graduate hiring falls from earlier peaks
Nasscom, the National Association of Software and Services Companies, is India's largest technology-industry representative body. Cognizant announced Nambiar's move to the association in August 2024. His warning reflects the concerns of an industry facing a changing business model; it does not measure graduate unemployment or establish how many jobs AI has removed.
The stakes extend beyond a single recruitment season. India's technology industry employs an estimated 5.95 million people in 2026, according to a report based on company disclosures and workforce interviews. Its established model of hiring graduates in bulk and developing them over time is under pressure.
Anthropic plans IPO marketing next week and a listing in the week of November 16, according to people familiar with the plans. Buyers are discussing valuations of $1.2 trillion–$1.3 trillion, with some weighing $1.1 trillion, as its filing reveals faster-growing compute costs and extensive infrastructure commitments.
Editorial Team
That report says Tata Consultancy Services, or TCS, hired more than 110,000 engineering graduates four years before fiscal 2026; its fiscal 2026 intake was less than half that level. Fellow IT services company Infosys recruited about 80% fewer fresh graduates in fiscal 2025 than three years earlier. These are company-specific comparisons over different periods, not an industry-wide measure of AI displacement.
A May 2026 hiring report offers a counterweight. Staffing firm Xpheno put the decline in lateral talent—experienced hires—across tier-1 IT services firms at 31,500 during fiscal 2026, while the firms maintained graduate intake commitments. The report said Cognizant had committed to 20,000 graduate hires alongside planned layoffs, and Infosys to 20,000 in fiscal 2027.
Industry experts cited in that account linked graduates' relative appeal to margin pressure and rising AI bills. Hiring cheaper beginners equipped with AI can be part of a cost-control strategy. Protecting graduate recruitment relative to experienced hiring can coexist with intake well below earlier peaks; neither pattern establishes that entry-level opportunities are secure.
The work that taught juniors is changing
The white paper How AI Is Reconfiguring Work in India explains the potential training problem. It comes from the Centre for Responsible AI, or CeRAI, an interdisciplinary research centre at the Indian Institute of Technology Madras, within its Wadhwani School of Data Science and AI.
Established in 2023 to research responsible AI development and deployment, the centre identified Google as its first platinum consortium member, contributing $1 million. Its launch announcement also described collaboration with Nasscom on responsible-AI course materials, training and toolkits. The academic paper and the industry warning therefore come from institutions with existing links.
The paper distinguishes tasks, jobs, occupations and aggregate employment outcomes. Automating one assignment does not make an occupation redundant. But when enough recurring work disappears, employers may redistribute the remaining duties among reviewers or supervisors, leaving too little to sustain a dedicated position.
In software, the authors describe AI absorbing routine code while people retain responsibility for architecture, complex debugging and technical leadership. Those senior responsibilities depend on judgment that juniors previously built through simpler assignments. Retaining expert work does not by itself preserve a way to acquire expertise.
The authors call their framework a structured hypothesis. It draws on literature, practitioner conversations and illustrative cases, rather than a representative labour-market sample. Predominantly formal-sector evidence cannot establish how widely each pattern occurs across India, and the paper does not systematically measure wages, work intensity or whether new roles absorb displaced workers.
Deployment economics also limit predictions of wholesale replacement. The paper argues that routine business-process outsourcing can use inexpensive, narrowly targeted systems, while more demanding applications face higher capability, integration and infrastructure costs. It cautions that subsidised AI pricing may not persist and describes customer-service automation being scaled back after quality problems. Technical capability alone does not establish a durable saving or an acceptable service.
CeRAI, IIT Madras, outlines seven pathways for job change and a continuum of human and AI involvement. Its framework is a structured hypothesis, not a representative labour-market measurement. Source: Centre for Responsible AI, IIT Madras.
AI assistance can strengthen learning too
Evidence from workers using AI challenges the assumption that automating parts of a job necessarily damages learning. In Generative AI at Work, researchers studied the staggered introduction of an AI conversational assistant using data from 5,172 customer-support agents. The November 2024 revision reports a 15% average productivity increase, measured as issues resolved per hour.
The distribution matters: less experienced and lower-skilled agents improved both speed and output quality. The most experienced and highest-skilled workers saw small speed gains and small quality declines. The researchers also found evidence that assistance facilitated learning.
This is evidence about people working with an assistant in customer support. It does not measure Indian graduate engineering recruitment or what happens when employers replace workers. It nevertheless demonstrates an alternative to losing the apprenticeship: AI can help beginners improve while they remain engaged in the work.
Employers are raising their expectations at entry. Mansee Singhal, country leader at Marsh Workforce & Rewards, which advises companies on hiring and talent management, said in the graduate-hiring report that applicants increasingly need technical depth and an understanding of business problems from the outset. Multinational companies' centres in India are also moving towards specialised work beyond routine back-office operations.
The companies say they are adapting. TCS and Infosys told the same report that their graduate programmes now include generative AI, emphasising hands-on learning at TCS and problem-solving and design thinking at Infosys. Kamal Karanth, co-founder of staffing firm Xpheno, cautioned that specialisation in roles such as machine-learning engineering and AI architecture still requires years of practical experience.
There is also a gap between saving time and using it to develop workers. Sumit Sarawgi, a technology-sector partner at management consultancy Oliver Wyman, said a company survey found only 12% of respondents in India reported employer plans for using time saved by AI; 48% reported no guidance. The account gives neither the sample size nor fieldwork dates; those shares cannot establish how common such planning is across Indian employers.
Who funds the next route into senior work?
The IIT Madras paper proposes redesigning entry roles around checking AI output, handling exceptions and evaluating performance. Its illustrative customer-support pathway runs from voice agent through escalation, quality evaluation and conversation design to operations supervision. It does not establish that employers are adopting such career paths widely.
Balaraman Ravindran, head of CeRAI and the Wadhwani School and a contributor to the paper, also led the centre at its 2023 launch. In an interview about the research, he said supervisory transitions require adequate training, suitable tools and safeguards, while acknowledging that research does not yet adequately answer how to make them work. He envisaged funding shared among the state, industry and learners, with the balance depending on the sector and the social impact of displacement.
Universities face an access problem as well as a curriculum problem. Academics interviewed for the graduate-hiring report called for real-world projects and assessments that test students' reasoning about AI-generated solutions, alongside wider mentorship, internships and faculty support beyond elite institutions. Moving more preparation into education will have limited reach if those opportunities remain concentrated.
New job categories offer possibilities, not guaranteed destinations for displaced graduates. An October 2025 government roadmap from NITI Aayog, developed with Nasscom and consultancy BCG, projects an opportunity for up to four million new jobs over five years as AI reshapes tech services. Its official release calls for a National AI Talent Mission spanning education and reskilling, alongside compute infrastructure and data access. This was a recommendation, not evidence that such a mission had delivered jobs. The IIT Madras paper stresses that the projection depends on coordinated investment and does not show whether displaced workers fill new roles.
The next decision is how employers and public training programmes fund specific routes into that work: which beginners enter, how long training takes, who supports them during it and whether earnings are preserved. Tracking their pay, workload and progression into expert roles would test whether redesigned junior jobs build experience as well as reduce current costs.
Internet scans and leaked databases expose the spread of ARTEX, a security-testing framework used in cyberattacks. Researchers documented largely autonomous compromises, but the records also show failed cashouts and uncertain counts of independent deployments.
Meta's strategy chief and former CFO is joining Fireworks AI ahead of a reported fundraising push targeting at least a $30 billion valuation. The inference company announced a $17.5 billion valuation in July, making the hire part of a much larger expansion than its older funding figures suggest.
Sand.ai says its $200 million Series A+ round lifts its valuation above $1 billion. Co-lead investor Wangsu plans inference and deployment cooperation, while its filing shows substantial losses behind the startup’s video AI ambitions.
An advocacy group estimates Interior Secretary Doug Burgum’s AI holdings at $18.1 million–$64.4 million as his department advances data centers on public land. His ethics agreement allows retained investments with recusals, while pending permits will determine both the application of those safeguards and the public’s return. Interior denies his decisions are driven by personal investments.
Microsoft is evaluating fewer neocloud partners while expanding its own infrastructure. Delivery-dependent contracts underpin supplier financing, while IREN reports rising revenue alongside a $6 billion funding target for its broader expansion.
Physical Superintelligence says its AI system found a proof of a conjecture limiting quantum query advantages on random inputs. Public proof files enable scrutiny, but do not establish the private system’s cost or repeatability.
OpenAI is expanding its relationship with Hacktron, whose researchers previously reported an attack chain into employee accounts, to test third-party connections with greater access to cybersecurity models.
Google has acquired Crusoe’s southeast Wyoming data center site, where it plans a long-term $17 billion investment. Crusoe describes completed pre-development work; independent reporting points to customer setbacks. Power delivery and carbon capture remain key tests.
The team behind document-parsing service SoMark has released OXR-1.0 code and weights. A benchmark retest supports a strong result, but speed and accuracy were measured separately, and users above funding or revenue thresholds need a commercial license.
OpenAI is reportedly developing a Business Hub with daily mention and click charts, product updates and business agents. The unannounced feature could give verified businesses direct visibility into ChatGPT responses, without revealing prompts or competitor rankings.
Palantir and Nvidia have specified an integrated stack for running AI in customer-controlled environments. The architecture builds on existing software and infrastructure offerings, with published configurations spanning 32 to 640 GPUs, while pricing and workload economics remain unproven in the launch materials.
Democrats’ incoming congressional majorities put hearings and investigations within reach, but AI rules and crypto legislation still face bipartisan bargaining and a Republican president.
Microsoft reportedly declined to renew Oracle server rentals as older chips became less useful to it. The decision comes despite continuing capacity shortages, while Oracle says demand for earlier-generation GPUs remains strong.
An investigation puts Nvidia’s future commitments and guarantees at $530.5 billion. Its disclosures show rapidly expanding supplier commitments and phased support for customers whose infrastructure could generate further chip sales, with cash payments depending on timing, demand and contract terms.
TSMC is reportedly planning a possible move into 3D DRAM at Nvidia’s request. The foundry already helps integrate AI processors with memory, but a manufacturing expansion has no disclosed factory or production date.
Physicl has raised a $10 million seed round and introduced Studio, a beta app for generating targeted physical AI data. Its reported customer gains concern 3D reconstruction, leaving broader robot reliability and commercial economics to be demonstrated.
New cloud partners in Southeast Asia must enroll in Nvidia’s Fleet Intelligence system, according to a report, linking preferred chip access to monitoring while leaving data sharing and existing partners’ obligations unresolved.
Airbnb sold intellectual property to a private AI lab controlled by CEO Brian Chesky in a $7.1 million arrangement that also reimburses expenses. It obtained a license to certain lab technology and future investment rights, but disclosed no asset valuation or detailed license terms.
Chip shares recovered part of Thursday’s losses in early Friday trading. Fears about data-centre construction face a test against AI-chip supplier Nvidia’s strong reported sales, growth forecast and conditional infrastructure-financing plans.
