# AI Labs Acquire Data Archives from Defunct Startups Failed companies are monetizing digital archives to satisfy AI training demands for real-world 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/8cebb6e7-db84-48d3-8723-94e466e8a81b ## Monetizing Failed Startup Archives AI labs are increasingly purchasing data from defunct startups to train their models. Keelan Miskell, who recently wound down his crypto venture, successfully sold his company’s codebase and workspace archives for significant profit. This emerging secondary market helps liquidators recover value from failed entities while providing labs with essential real-world records. ## Limits of Synthetic Datasets The demand for proprietary data has surged as training labs encounter performance ceilings with public and synthetic datasets. Researchers now prioritize real-world records to improve the reasoning capabilities of AI agents. This shift forces companies to seek unconventional sources, turning the digital remnants of failed businesses into highly valuable training commodities. ## Ethical and Legal Hurdles The practice of selling startup data to AI labs raises significant legal and ethical concerns regarding user privacy and intellectual property. While liquidators view these archives as assets, the original intent of the data collection often did not include AI training. Regulators are now scrutinizing whether such sales violate existing data protection agreements. ## Advancements in Interspecies Communication Researchers are leveraging AI to decode complex animal communication, including howls, squeaks, and clicks. By analyzing these vocalizations, scientists hope to gain insights into the inner lives of various species. New hardware and machine learning algorithms have made it easier to capture and process these signals, though interpreting the underlying meaning remains difficult. ## Challenges of Lunar Colonization NASA and international partners are racing to establish lunar bases, but the moon remains a hostile environment for human habitation. Future lunar cities must withstand extreme temperature swings, cosmic radiation, and razor-sharp dust. Success depends on autonomous robotics, nuclear power integration, and the ability to utilize local lunar resources for long-term survival. ## Cybersecurity and Election Integrity Recent disruptions to the US Federal Election Commission website highlight growing vulnerabilities in critical infrastructure. As midterms proceed, software flaws in voting machines and the spread of AI-generated misinformation on social media platforms threaten democratic processes. Experts warn that internet-wide vulnerabilities are being identified by AI tools faster than developers can patch them. ## Corporate Sustainability Controversies Apple’s strategy to achieve net-zero emissions by 2030 involves planting millions of eucalyptus trees in Brazil. While the company frames these industrial tree farms as essential carbon offsets, critics argue the initiative may cause ecological harm. This backlash underscores the tension between corporate carbon-neutral goals and the actual environmental impact of large-scale offset projects. ## Strategic Shifts in Technology The tech landscape is evolving rapidly, from the commercialization of failed startup data to the development of Martian rocket fuel catalysts. As billionaires fund campaigns against wealth taxes and Gen Z pivots toward digital detox tools, the industry faces mounting pressure to balance innovation with ethical responsibility, political stability, and long-term environmental sustainability. ## 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._