# AI Labs Acquire Data Archives from Defunct Startups Failed companies are monetizing digital archives to satisfy the growing demand for training data. - **Source:** [MIT Technology Review](https://www.technologyreview.com) - **Published:** 2026-10-28 - **Topic:** [Data & AI](https://brifing.si/c/data-ai) - **Original article:** - **On Brifing:** https://brifing.si/s/0bc592f3-34d8-497d-8732-4155f7f8eaa7 ## Monetizing Failed Startup Archives AI laboratories are increasingly purchasing data from defunct startups to train their models. Keelan Miskell, a former crypto startup founder, successfully sold his company's codebase and workspace archives for tens of thousands of dollars. This emerging market helps liquidators recover value from failed ventures while providing AI developers with essential real-world records. ## Data Scarcity Drives Demand The demand for high-quality training data has surged as AI labs hit the limits of existing public and synthetic datasets. Developers now prioritize real-world records to improve the performance of AI agents. This shift forces labs to look beyond traditional sources, turning the remains of failed startups into valuable assets for machine learning. ## Ethical and Legal Implications The practice of selling startup data raises significant legal and ethical questions regarding privacy and ownership. While founders may seek to recoup losses, the secondary use of proprietary codebase and internal workspace data remains largely unregulated. Industry observers are now debating the long-term implications of treating failed company archives as commodities for AI. ## Advancing Interspecies Communication Researchers are leveraging AI and specialized hardware to decode animal communication patterns. By analyzing how species like birds and marine life interact, scientists hope to gain insights into their inner lives. While devices can now capture complex howls and clicks, interpreting these signals remains a significant technical challenge for current machine learning models. ## Lunar Base Construction Challenges NASA and China are competing to establish lunar bases by the end of the century. The moon presents a hostile environment characterized by cosmic radiation, razor-sharp dust, and extreme temperature swings. Success depends on developing autonomous robots and nuclear power systems to ensure the base remains self-sustaining despite the harsh, volatile lunar conditions. ## Political Obstacles to Space Beyond technical hurdles, the chief obstacle to establishing a lunar city is political will. Building a permanent human presence requires sustained international commitment and massive resource allocation. As nations race to secure strategic positions on the moon, the feasibility of these projects continues to hinge on long-term geopolitical stability and funding priorities. ## AI Vulnerability and Security Matthew Prince, CEO of Cloudflare, warns that AI is identifying software vulnerabilities faster than developers can patch them. He predicts a persistent state of internet-wide security threats for the next two years. This rapid discovery cycle forces cybersecurity teams to accelerate their response times to mitigate potential exploits across critical digital infrastructure. ## Corporate Carbon Neutrality Debates Apple is pursuing a net-zero goal by 2030 through massive tree-planting initiatives in Brazil. Critics argue that these industrial eucalyptus farms may cause more environmental harm than good. This controversy highlights the tension between corporate offsetting strategies and the actual ecological impact of large-scale reforestation projects intended to balance carbon footprints. ## Sources covering this story - https://www.technologyreview.com/2026/10/28/1147767/the-download-dead-startups-fueling-ai-interspecies-communication/ — --- _Summary written by Brifing from the original reporting. Copyright in the original article belongs to its publisher; read the full piece at the link above._