Make or Buy: Standardtool, Plattform oder Eigenentwicklung? (KI-Strategiebriefing am 05.06.2026)

Here's what you can expect this week: 1. The most relevant AI developments of the week: Anthropic and OpenAI before their IPOs: Two of the most important AI labs on their way to trillion-dollar valuations AI between protection and vulnerability: Security gaps in critical infrastructure, meta-chatbot hack, and new risks from agent privileges OpenAI Codex becomes a universal app: Bankers, analysts, designers, and other specialists are increasingly using a tool originally intended for developers 2. This week's focus: Make or buy: Standard tool, platform, or in-house development? When is ChatGPT or Copilot sufficient? When is a central AI platform necessary? And when does in-house development make sense? Many companies have licensed and tested initial AI tools. But after a few months, it often becomes clear: More licenses don't necessarily mean real results. Data silos, shadow AI, inconsistent processes, and rising costs emerge—but often no measurable added value. ``` This deep dive therefore focuses on the central architectural question of AI transformation: Which use cases belong in standard tools, which require a platform with centralized knowledge, rights, and compliance – and where can in-house development create genuine strategic differentiation? The key message: The AI ​​decision of the future is no longer simply a question of tools. It's about thoroughly evaluating use cases, organizing data and processes, and building an AI landscape that is scalable, secure, and effective in the long term.