In October 2025, Cambridge-based AI biotech startup Ignota Labs acquired all remaining clinical assets from defunct US cancer developer Kronos Bio, launching an initiative to resurrect clinically failed drugs using artificial intelligence. Kronos Bio, once valued at $3.5 billion during its IPO, was ultimately acquired for $0.57 per share after multiple candidate programs stalled in phase II trials due to safety concerns and suboptimal clinical positioning. Ignota Labs will leverage its proprietary AI platform, SAFEPATH, to diagnose the precise molecular and biological reasons for these failures and execute targeted structural modifications. Traditional drug repurposing simply assigns existing molecules to new indications without altering their chemical structure. Ignota’s approach focuses on repairing flawed candidates. SAFEPATH integrates multimodal datasets, including molecular architectures, protein-binding affinities, gene expression profiles, preclinical records, and clinical adverse events. The platform identifies cross-layered patterns invisible to conventional analysis, such as off-target binding, abnormal metabolic clearance, or unintended tissue distribution. Upon pinpointing the failure mechanism, SAFEPATH generates optimized molecular editing proposals that undergo virtual screening and experimental validation before advancing to clinical development. The acquired portfolio centers on three high-value candidates: the CDK9 inhibitor istisociclib, alongside the SYK inhibitors entospletinib and lanraplenib. Ignota will prioritize the SYK inhibitors for immune thrombocytopenia and chronic lymphocytic leukemia, targeting patients who have developed resistance to standard therapies or suffer from unacceptable toxicity with current treatments. Because the original mechanisms proved biologically sound, correcting the safety and positioning deficits could unlock markets valued at nearly $100 billion globally. Ignota Labs was founded in 2021 by Layla Hosseini-Gerami, Jordan Lane, and Sam Windsor. The team combines expertise from BenevolentAI, Merck, and Google DeepMind’s AlphaFold initiative. Following a $6.9 million seed round in February 2025, the company expanded from four to fourteen researchers by October, integrating machine learning, structural biology, medicinal chemistry, and clinical translation. Despite this progress, Ignota co-founder Hosseini-Gerami emphasizes a systemic bottleneck: decades of failed clinical trial data remain trapped in proprietary corporate servers. The pharmaceutical industry historically prioritizes successful publications, leaving a vast, unutilized archive of toxicity and efficacy boundaries. Ignota argues that establishing secure data-sharing frameworks and reciprocal exchange models could transform failed experiments into high-yield assets. By applying machine learning to previously discarded candidates, the company aims to prove that fixing broken science is as commercially viable as discovering entirely new molecules, potentially reshaping how the industry manages risk and accelerates therapeutic development.