Fulcrum Insights
AI News
A daily AI briefing for MSPs and IT leaders: the stories that matter, what to do about them, and the sources behind every item.
Latest briefing
Issue No. 001Security · Open Models · AI Coding Tools · Compliance
A poisoned Claude Code extension, Mistral's open trillion-parameter model, and US-only inference
A malicious Claude Code extension reached up to 5,000 Cursor users with an SSH backdoor, Mistral released open weights for its trillion-parameter Large 4, and providers shipped new data-residency and routing options worth checking.
Story 1 of 3