February 2026

February's AI coverage spans an increasingly uneasy landscape where technical advances collide with governance, security, and cultural consequences. Coding agents and self-improving models dominate the innovation narrative — from Anthropic's viral Claude Code to frameworks for managing AI context and agent systems — while a parallel wave of safety discourse examines sabotage risks, prompt injection attacks, shadow AI, and exploitable enterprise agents from Microsoft and ServiceNow. Broader social and institutional friction runs throughout: OpenAI contends with employee dismissals, ethical comparisons to Facebook's trajectory, and an ongoing dispute between OpenClaw and OpenAI; meanwhile psychologists flag psychosis-adjacent behaviors in chatbot users, creators form coalitions against AI encroachment, and commentators debate whether AGI is illusion or imminence. Underlying it all is a design and purpose crisis — questions about what human creativity, software development, and even identity mean in a world where AI increasingly does the work.