Weigh the opportunities against the regulatory and ethical problems of using AI in drug discovery, with particular reference to rare and neglected diseases in India.

AI-led drug discovery uses machine learning to find, screen and refine molecules, shortening the slowest stage of research. The Technology Development Board's (TDB) ₹86 crore support to Bengaluru's Peptris aims to take four AI-discovered candidates, including one for Duchenne Muscular Dystrophy (DMD), "from bench to clinic" [1]. It shows both the promise of AI and the gaps in how it is governed.

Opportunities

• Faster, cheaper discovery: AI finds promising molecules and disease mechanisms quickly [1]. This matters for rare diseases, whose small markets put off large pharma companies.
• Unmet need: the VIGOR programme targets conditions with few treatments, such as DMD and neuropathy [1]. Domestic drugs could ease costly imports under the National Policy for Rare Diseases, 2021.
• Closing the funding gap: the ₹1 lakh crore RDI Scheme funds projects at TRL-4 and above through loans and equity [2]. Private firms pay for only about 36% of India's R&D spending [3].

Regulatory problems

• Rules lag behind: WHO notes that most countries have no specific laws on AI in health [4]. The New Drugs and Clinical Trials Rules, 2019 do not set out how CDSCO should assess AI-generated evidence.
• The bench-to-clinic gap remains: AI hits still need full preclinical and IND-enabling studies [1]. Rare diseases have small patient pools, which makes trials hard to run.
• Offshore execution: the work runs through global CRO networks [1], so skills and spending may go abroad.

Ethical problems

• Bias and data rights: models trained on non-Indian data may miss Indian genetic diversity. Consent and data ownership are also unclear. WHO lists inclusiveness and equity as core principles [4].
• Opacity and accountability: "black-box" predictions make it hard to say who is responsible when they are wrong [4].
• Access: half the project is public money, yet the release sets no pricing or access terms. It also funds an alopecia drug alongside DMD [1].

On balance, the opportunities outweigh the risks, especially for neglected patients the market ignores. To secure them, CDSCO should issue guidance on AI evidence in IND filings, RDI deals should include access clauses for rare-disease drugs, and India should build representative health datasets. Together these would align AI-led drug discovery with the right to health under Article 47 and SDG 3.