🤖AI Newsletter
1. September 2026 · 04:45 Uhr
1Bank of England Chief Warns G20 of AI Investment Bubble and Cyber Risks from Frontier AI
THE DECODER The Bank of England chief warns of an AI investment bubble driven by overvalued companies and rising leverage, which poses systemic risks during market corrections. Additionally, cyber risks from advanced AI models and interconnected dependencies between AI firms and hyperscalers are identified as stability risks. This could lead to stricter regulation and increased compliance requirements for AI investments.
2OpenAI Reports Billion-Dollar Milestone in Advertising Business Ahead of IPO
THE DECODER OpenAI has scaled its advertising business to an annualized revenue run rate of $1 billion, signaling substantial growth beyond its core business ahead of a planned IPO. This diversifies revenue streams and strengthens the valuation foundation for the IPO, while intensifying competition in ad-supported AI monetization.
3EU Commission Classifies ChatGPT as Search Engine for First Time
THE DECODER The EU Commission classifies ChatGPT for the first time as a "very large search engine" under the Digital Services Act, requiring OpenAI to undertake extensive compliance measures (risk assessments, transparency reports) by 2026. This marks the first regulatory equalization of AI chatbots with traditional search engines and sets new standards for European AI governance. The classification could serve as a blueprint for further regulations and increases operating costs and market entry barriers for AI providers.
4OpenAI Apparently Experiments with Payment Only for Successful AI Work
THE DECODER OpenAI and other AI providers are experimenting with outcome-based pricing instead of traditional subscriptions – payment occurs only upon demonstrated success of the AI solution. This signals a paradigm shift in the SaaS market and addresses the central trust problem enterprise customers face with AI investments. The shift could pressure revenue models of major providers, but also creates competitive advantage through risk mitigation for customers.
5Mac mini Becomes AI Training Computer: OpenAI and Anthropic Massively Deploy Apple Hardware
THE DECODER AI market leaders like OpenAI and Anthropic are purchasing Apple hardware (Mac mini/Studio) in massive volumes for training agents, significantly boosting Apple's Mac segment and driving demand so strongly that high-end models are sold out. This signals a new hardware use case beyond classic end-user applications and creates an additional high-volume revenue stream for Apple in the B2B AI sector.
Tokens: 1,607(996 in · 611 out)