Internal research conducted at Anthropic has revealed that since the dawn of 2025, there has been a marked decrease in the proportion of entry-level roles within its technical division, along with a significant drop in the percentage of newly recruited entry-level staff. By December 2025, the availability of entry-level job vacancies had diminished considerably, whereas the corresponding proportions in non-technical departments stayed constant. A key driver behind this transformation is the advancement in AI agent capabilities. By August 2026, Anthropic's internal AI agents were clocking in working hours six times longer than their human counterparts, with an astonishing 19-fold surge observed in the research and development sector.

Tasks such as coding and bug detection are now predominantly managed by AI tools like Claude, resulting in a particularly sharp decrease in the hiring of entry-level employees for roles heavily influenced by AI. The widespread integration of AI tools has substantially amplified work productivity across diverse departments. Nevertheless, Anthropic also cautions that the study's outcomes may not accurately mirror the broader economic landscape, as swift business expansion and employees' proactive adoption of AI could also skew the data.

A more profound worry is that the dwindling number of entry-level positions could restrict the avenues for nurturing professional expertise, a situation dubbed the 'tragedy of the cognitive commons.' This implies that while firms boost efficiency by substituting entry-level roles with AI, it might undermine the future pipeline of senior talent in the sector. The ramifications of this shift may only start to surface gradually between 2030 and 2045, and the enduring impact of AI on employment frameworks and vocational training systems still necessitates verification through more extensive economic data.