Telangana's AI Framework, AI Mission, AI Strategy and AI City (2020-2026)

Telangana's AI programme is run by the Emerging Technologies Wing of its ITE&C Department. The department's AI page says the Government of Telangana officially declared 2020 as the Year of AI, set up Telangana's AI Mission (T-AIM) with NASSCOM, and issued an AI Framework with six pillars. It also lists an Applied AI Research Centre focused on healthcare and mobility (with Intel, IIIT Hyderabad and the Public Health Foundation of India as co-founding partners) and a virtual Centre for Responsible Deployment of Emerging Technologies, set up with C4IR India of the World Economic Forum and initially focused on agriculture and law enforcement.⁴

The state's 2024 strategy cites earlier technology policies as foundations: the Telangana Open Data Policy (2016), the Blockchain Policy (2019) and the ICT Policy (2021-26), as well as an ecosystem of incubators led by T-Hub and a high concentration of AI patents.² The programme therefore sits in policy documents, not legislation, and the state has not enacted a dedicated AI law.

The AI Framework, dated June 2020, states a vision of establishing Telangana as a global hub for AI and sets a target of AI contributing USD 5 billion to the state's IT exports by 2025 while ensuring that products and services are deployed responsibly and ethically. It identifies focus areas of healthcare, agriculture, mobility, education and law enforcement, and records that the state had assembled a calendar of 74 Year of AI events with 12 ecosystem partners.¹

The six pillars are: (1) a Data Exchange Platform and High-Performance AI Computing facility, building on the 2016 Open Data Policy and an open data portal then hosting 192 datasets, with data-provider certification; (2) Skilling and Education, with a target of at least 30,000 additional AI workers by the end of 2022; (3) Governance, Ethics and Privacy; (4) Research and Innovation; (5) Enabling Adoption and Community Collaboration; and (6) an AI-Innovestment Fund for AI for social good, including in law enforcement.¹

The governance pillar is the most relevant to AI regulation. It promises a working committee, the AI Ethics Group, led by academics and including industry and research bodies such as the WEF and NASSCOM, to identify concerns of accountability, bias, transparency, privacy, security and safety and to release advisory guidelines, codes of practice and governance frameworks. Government AI procurement guidelines are to encourage explainable AI and mechanisms for interpretability, and the framework asks officers to know where human review is required for decisions with legal effects on individuals. Companies using government data are to file compliance reports on ethical use, state data governance guidelines are to be aligned with the then-pending Personal Data Protection Bill, 2019, and employees of IT companies using state datasets would take a mandatory data ethics course.¹ The framework also lists incentives, including SGST reimbursement for small AI start-ups and subsidies for patent filing.¹

The state released its AI-Powered Telangana strategy document and implementation roadmap at the Global AI Summit held on 5-6 September 2024 at the Hyderabad International Convention Centre. The document commits the government to a three-year plan for 2024-2027 and identifies six pillars.²³

• AI Direction Setting. A Telangana AI Advisory Council working with the Chief Minister's Office, described as the "NITI Aayog for AI in Telangana", is to set direction on research, public policy, talent, ethical and responsible adoption and the AI City, supported by an AI Research & Collaboration Network on a hub-and-spoke model. The Emerging Technologies Wing implements the plan and every department is to appoint an AI nodal officer.²
• AI Equipped Government. Departments are to move from problem identification and proofs of concept to pilots and tendered scale-up, with an aim to raise officials' productivity by over 20%.²
• AI Ready Local Datasets & Models. Government datasets are benchmarked from L1 (catalogued but non-digital) to L3 (standardised, citizen-level data in compliance with the Digital Personal Data Protection Act). The state plans Telugu language models, 50 data annotation centres expected to create 10,000 jobs, a Telangana Vaibhavam AI programme aiming at 10 billion tokens of cultural data, and the TGDeX data exchange with public dashboards.²
• AI Skilled Population. The state aims to train 5 lakh professionals by 2027, to introduce AI curricula in senior secondary schools and to run an AI literacy campaign.²
• AI Enhanced Ecosystem. Subsidised compute for government departments and selected early-stage start-ups through hyperscalers, with unified AI services.²
• Future Ready AI City. A 200-acre AI City hosting global corporations, start-ups, investors, research centres, data centres and high-performance computing, with a 'Centre for the Future' and district-level AI nodes.²

A section on generative AI calls for guidelines and frameworks for evaluating AI fairness and responsible AI implementation, but, unlike Tamil Nadu's policy, it does not set out a scoring or approval regime for public-sector AI.² MediaNama's summary of the strategy reported the same six elements under slightly different labels.⁵

Progress reporting comes mainly from the state's own documents. The July 2025 TGDeX paper, written by the ITE&C Department with JICA, says the state has run more than 30 department pilots, including a driver-assistance system on TGSRTC buses under Project iRASTE that it says had a 40% lower rate of fatal crashes than buses without the system between April 2023 and March 2024, and MITRA-TG, an anti-narcotics chatbot piloted in more than 1,000 schools. It reports that T-AIM's RevvUp programme and the NUage programme of T-Hub's AI/ML centre MATH had supported over 200 AI start-ups, that MoUs had been signed with Yotta, CDAC and NTT Data for compute, that data annotation centres were being set up with Uber in Karimnagar and Warangal, and that the state was working towards an AI University to train 200,000 AI engineers.³ These figures have not been independently verified for this entry.

TGDeX. The Telangana Data Exchange builds on the state's Agriculture Data Exchange (ADeX, 2023). The paper describes it as India's first state-led digital public infrastructure for AI, offering curated public datasets, reusable models, open-source tools and GPU compute in a sandbox, with role-based access control and compliance with data protection law, and as complementary to the IndiaAI Mission's AIKosh. In April 2025 the Telangana AI Rising Grand Challenge tested the model: departments shared datasets on six problem statements and over 240 start-ups submitted more than 420 proposals. Over five years TGDeX aims to host 2,000+ AI-ready datasets, engage 3,000+ start-ups and support 50+ full-scale deployments in the state, and to improve service delivery for over 1 crore citizens by 2027.³

Infrastructure investment. On 5 September 2026 the state announced with Tata Consultancy Services' HyperVault subsidiary a data-centre campus of up to Rs 70,000 crore on 264 acres at the Future City, with capacity of up to 1 GW and a target inauguration date of 2 June 2028; MediaNama noted earlier projects with CtrlS (Rs 10,000 crore) and NTT Data/Neysa (Rs 10,500 crore). The report does not describe the AI City itself.⁹

The state's AI plans have been scrutinised mainly through its identity and welfare systems rather than through the strategy document.

RTDAI facial authentication. The Real Time Digital Authentication of Identity system uses AI liveness detection and face matching and is used for pensioner life certificates, learner's licences, degree admissions on the DOST portal and an e-voting pilot. In September 2025 researcher Srinivas Kodali complained to the Telangana Chief Electoral Officer that voter photographs had been shared with the state and repurposed for the system, arguing this was unlawful under the Representation of the People Act, 1950 and the Puttaswamy privacy judgment, and MediaNama reported that accuracy figures in the state's own presentation varied widely across gender and age groups.⁶⁷ On 17 October 2025 MediaNama reported that the Chief Electoral Officer confirmed sharing EPIC data through API access for the Pensioners' Life Certificate Service while denying sharing for any facial recognition system.⁷

Unified Card. In June 2026 the state announced plans, as MediaNama reported from a government press release, for a 'Unified Card' that would aggregate data from welfare departments and a socio-economic survey in an AI-based, 360-degree beneficiary profiling system. Activist SQ Masood called it permanent state surveillance infrastructure, and MediaNama's own questions concerned data scope relative to the Aadhaar Act, reuse without fresh consent and whether AI would only flag duplicates or make decisions. The report records no government response.⁸

The strategy sections reviewed for this entry do not describe an independent oversight or grievance mechanism for such systems, and the 2020 framework's promised AI Ethics Group is not shown in the sources reviewed to have issued public guidelines.¹²

As of 15 October 2026 the AI Framework (2020) and the AI-Powered Telangana strategy (2024-27) are the operative documents, and the ITE&C Department continues to host them. The strategy period ends in 2027, and the headline 2027 skilling and TGDeX service-delivery targets had not been reported as met in the sources reviewed. I could not verify that the AI Advisory Council has been formally constituted, that the AI City has begun construction, or that the AI Ethics Group promised in 2020 has published guidelines. No Telangana AI statute or separate AI policy dated 2025-2026 was found.¹²⁴