Swedish government agencies to ban ChatGPT and Claude
claude copilot microsoft
| Source: HN | Original article
Swedish government offices have barred ChatGPT and Claude from civil servant computers, leaving Microsoft Copilot as the sole AI tool permitted.
Swedish government offices will block direct access to OpenAI’s ChatGPT and Anthropic’s Claude on civil‑servant computers, effective Monday, October 26. The decision, announced on October 23, leaves Microsoft Copilot as the sole AI tool officially approved for everyday use. Staff may still reach ChatGPT or Claude through a tightly controlled internal platform, but the default, unrestricted route is being shut down.
The ban reflects growing concerns over data security and compliance in public administration. Officials cited the need to safeguard sensitive government information and to ensure that any AI service meets stringent Swedish and EU privacy standards. By channeling users toward Copilot—already vetted and integrated with Microsoft’s security ecosystem—the government aims to reduce the risk of accidental data leakage or misuse of unverified generative models.
The move underscores a broader tension between rapid AI adoption and regulatory caution. While other public bodies in Europe are experimenting with a range of agents, Sweden’s hard line may set a precedent for how state institutions manage third‑party AI services. It also gives Microsoft a de‑facto monopoly on approved generative AI within the Swedish bureaucracy, potentially influencing procurement decisions in other Nordic ministries.
Watch for reactions from OpenAI and Anthropic, which may seek to address the government’s security concerns, and for any follow‑up guidance from Sweden’s data‑protection authority. Parallel developments in neighboring countries could either reinforce a regional trend toward tighter AI controls or prompt a counter‑push for more open access, shaping the future landscape of public‑sector AI use across the Nordics.
Sources
Back to AIPULSEN