Anyone else just wondering every week why AI hasn't taken their job yet?

A viral Reddit post from a UK-based law firm employee is sparking discussion about the real-world pace of AI-driven job displacement. The employee, in an entry-level administrative role, details tasks perfectly suited for automation: responding to templated emails, filling Word documents from PDFs, and negotiating settlements using repetitive arguments. They note these are 'a piece of cake for AI' and recall a lawyer predicting years ago that AI could already perform core legal functions, lacking only trust and indemnity insurance.

The poster reveals their firm is actively developing internal AI software to process opposing solicitors' claim documents and automate outputs, yet they've observed 'zero impact' on staffing, even for simple admin roles. This contrasts with headlines of job losses in more complex tech sectors. The core question posed is why adoption is so slow for seemingly automatable tasks.

The discussion points to non-technical barriers as the primary bottleneck. Legal and regulatory frameworks, especially in jurisdictions like the UK, require robust solutions for liability, data privacy (handling sensitive client information), and professional indemnity before AI can assume responsibility. Furthermore, enterprise integration—rewiring legacy workflows, ensuring reliability, and managing change—is often slower than developing the AI capability itself. The post underscores that the timeline for AI workforce impact is less about technical feasibility and more about the complex ecosystem of trust, regulation, and business process transformation.