Between 2004 and 2023, Hindustan Aeronautics Limited built 222 Su-30MKI fighters at Nashik under license from Russia. It is among the largest licensed-production runs of a modern combat aircraft undertaken outside the country that designed it. Two decades of that work have not given India unrestricted design authority over the aircraft. The Su-30MKI is not an isolated case. India has assembled fighters, tanks, submarines, helicopters and missiles under transfer-of-technology arrangements for six decades, and the same pattern recurs.
This report asks why. It argues that much of the explanation lies in how India has understood technology transfer. The system treats it as something a contract can deliver, and Indian procurement has become steadily better at specifying what is to be handed over. A contract cannot ensure that Indian engineers understand the technology, that the organization keeps the people who learned it, that the knowledge is improved on, or that it is used again in the next program. The distinction that matters is between receiving technology and absorbing it.
To make that distinction usable, the report sets out a ladder of technological depth. It runs from assembling imported kits, through licensed manufacture and mastery of production processes, to design participation and, at the top, design authority: the right and technical ability to change a design without the originator’s permission or assistance. Production depth and technological depth are separate axes. A program can climb one a long way without climbing the other. The report then sorts the conditions for absorption into three groups. Some are directly contractible: documentation scope, license terms, tooling, training hours, indigenous-content thresholds, IP and source-code rights, penalties. Some can be specified but not guaranteed: training quality, export releasability, supplier-ecosystem development, liaison mechanisms. The rest sit outside a contract altogether, among them the recipient’s prior technological baseline, the absorption of tacit knowledge, workforce retention across programs lasting decades, the supplier’s incentive not to create a future competitor, and institutional continuity. Most of what decides the outcome lies in the last two groups.
The policy record, read across nine procurement procedures from DPP 2002 to the draft DAP 2026, shows three things. Transfer of technology is never clearly defined in twenty-four years of procedure, and functions as a label in contracts rather than as a technical category. Indigenous content rose without design control. And no institution is made responsible for absorption, so procurement rules improved while learning outcomes did not.
The audit record supports this. The Comptroller and Auditor General’s performance audit of defense offsets covering 2007 to March 2018, Report No. 20 of 2019, examined 46 contracts carrying obligations of ₹66,427 crore (about USD 7.62 billion). Of the ₹19,223 crore (about USD 2.21 billion) that had fallen due by December 2018, vendors claimed ₹11,396 crore (about USD 1.31 billion), and the government accepted ₹5,457 crore (about USD 626 million). The audit found no case in which a foreign vendor transferred high technology to Indian industry.
Four programs are then scored on a five-stage chain: access, transfer, assimilation, reapplication and authority. The T-90S Bhishma shows deep assimilation, with AVNL stating in 2026 that the 1000th India-built tank had been delivered from Avadi at around 80 percent indigenous content, alongside little evidence that two decades of that work built a broader Indian tank-design ecosystem. The Su-30MKI shows the trade made explicit: a 2006 schedule compression cut fully indigenized Phase IV aircraft from 114 to 60 and raised fully imported Phase I and II aircraft from 8 to 34, while the project cost rose from ₹22,122.78 crore to ₹39,605.95 crore, an increase of 79 percent. Against that, HAL integrated the 2.5-tonne BrahMos onto the same airframe without the original designer, which is the real integration authority. The K9 Vajra-T traveled furthest short of design participation: more than 80 percent indigenous work packages, and a Hazira industrial base that went on to feed the DRDO and L&T Zorawar prototype, with design authority still resting with Hanwha. BrahMos climbed highest and is the least representative, because it was a joint development rather than a transfer.
Five findings follow. Technology transfer is not one thing, and Indian policy has used the term without separating its levels. More local production has not meant more technological capability. Many of the factors that decide whether technology is absorbed lie outside the contract. The weakness appears after production begins, where knowledge stays tied to the line, the factory or the program that acquired it. And that failure to travel persists because no authority is tasked with making it travel. The report is diagnostic, and a later report in the series carries the detailed recommendations. Four implications follow directly from the findings. Someone has to be responsible for absorption, connecting the organizations that buy, produce, and design. India should measure absorption rather than localization, asking what its engineers can now do independently and which technologies have moved into other programs. Where the objective is technological capability rather than domestic manufacture, the evidence better supports deeper participation in development than deeper transfer. And,
most cheaply of all, the procurement procedure should stop using transfer of technology as a single undifferentiated term, so that a contract can state which level it is buying and an auditor can say whether it was delivered.