Portugal
Humans, the startup that prevents AI agents from “getting crazy”, wins Portugal’s Web Summit contest
Publication: ECO
Author: Mariana Bandeira
Source published: 12/11/2026, 13:25:55 UTC
Briefing created: 22/11/2026, 19:48:21 UTC
Based on a feed excerpt; full article content has not been acquired. Source evidence retrieved 12/11/2026, 13:56:25 UTC.
Briefings
Portuguese startup Humans won the main prize and 'Best Performance' award at the 2026 Web Summit in Lisbon, securing a 15,000 euro prize. The company, founded in Madeira in 2025 by five founders, develops risk control systems for autonomous AI agents to prevent them from becoming uncontrollable. The startup faced a critical challenge during the event where a Ciberataque autónomo da OpenAI caused a shutdown due to a patient's depression, requiring the startup's team to work all day to complete pitches. The startup's CEO, Pedro Andrade, explained that their business focuses on measuring and controlling AI agent risks before implementation, aiming to prove to clients that their agents are under control.
Key points
- Humans secured a 15,000 euro prize and the 'Startup com Melhor Desempenho' award at the 2026 Web Summit in Lisbon.
- The startup was founded in 2025 in Madeira by five founders and currently operates from Sao Francisco.
Context
Humans competes against 78 startups in the 2026 Web Summit Road 2, organized by Startup Portugal and Galp Energia. The event aims to recognize Portuguese innovation for internationalization goals. The startup's success highlights the growing importance of AI agent risk management in the tech sector.
Limitations
Evidence covers only the main prize and 'Best Performance' award; no details on other categories or specific technical features of the risk control system are provided. The startup's future product launch in January is mentioned but no specific details are given. The event's impact on the broader AI landscape is not quantified.
Generated by Bittrees-hosted qwen35-2b sha256:aaf42c8b7c3c on 22/11/2026, 19:48:21 UTC. Based on publisher-provided feed evidence; may not cover the full original article. Check the source for details.
IPFS archive pending.