"Next year, tech companies will basically only hire AI talent," predicts Lin Fan, founder and CEO of professional networking platform Maimai. He clarifies this isn't limited to technical personnel; rather, companies will place increasing value on AI proficiency across all job functions and positions.

Data from the Maimai executive talent think tank report, "Agents Reshaping the Workplace — AI Talent Flow Report," shows that from January to July 2026, new AI-related job postings in the new economy sector on the Maimai platform surged by 789.47% year-on-year. AI is rapidly becoming a baseline capability requirement for roles, with 34.09% of surveyed professionals reporting that their employers have implemented management measures around employee AI skills — including setting AI thresholds in hiring, incorporating AI into performance reviews, and organizing reassignment or retraining assessments.

According to the report, the average monthly salary for new AI positions during January to July 2026 reached 63,160 yuan, up 2.74% from 61,474 yuan during the same period last year. By city, Beijing, Shanghai, and Shenzhen ranked as the top three for new AI job postings. Among the top ten cities for AI postings, Hefei, Chengdu, and Suzhou saw the fastest growth, with year-on-year increases of 1,570.37%, 1,158.70%, and 1,068.18%, respectively.

AI recruitment is currently driven by large and medium-sized enterprises while simultaneously expanding into smaller organizations. Companies with more than 10,000 employees contributed 33.49% of all new AI positions, while firms with fewer than 100 employees saw a 1,040.57% year-on-year increase in AI postings — the highest growth rate among companies of all sizes.

Lin emphasizes that AI talent shouldn't be pigeonholed as foundational large-model researchers. "Talent at the foundational level earning ten million yuan annually is beyond the reach of most companies. What most businesses actually need are people who can build Agent applications and those who can integrate Agent systems into core business workflows."

The report categorizes AI talent and roles into three tiers: AI technical infrastructure, AI application development, and AI business efficiency. From January to July 2026, AI application development roles accounted for 60.97% of all AI positions, AI technical infrastructure roles made up 24.55%, and AI business efficiency roles represented 14.48%. The focus of AI hiring is shifting away from model training and research toward application building, on-site delivery, and operational efficiency improvements. Among the most in-demand application development roles, Agent development engineers top the list.

The salary premium associated with AI capabilities is now emerging beyond purely technical positions. During January to July 2026, among the ten major job categories in the new economy, sales roles with AI skills commanded the highest premium at 27.80%, with an average monthly salary of 42,817 yuan. Operations roles with AI proficiency saw a 20.30% premium, averaging 48,120 yuan per month, while HR and headhunting positions with AI skills carried a 17.50% premium, averaging 38,168 yuan monthly.

Lin believes that AI proficiency is becoming a fundamental workplace skill, and compensation is increasingly flowing toward those who can leverage AI to drive better business outcomes. He offers a straightforward benchmark for assessing AI capability: "Can this person make an Agent work autonomously for more than one hour?" In his view, achieving this demonstrates that the user isn't just calling tools but can define objectives, design constraints, and close the validation loop with AI.

According to the report, in August 2026, 27.01% of respondents reported their longest single-session Agent run lasted over one hour, with 16.59% exceeding two hours. Among those who ran coding Agents continuously for over an hour, 63.30% said AI now handles more than half of their work, and 55.80% attributed output gains of over 50% to AI. In contrast, among other groups, these figures stood at 20.50% and 15.80%, respectively.

Looking ahead, Lin predicts that Agents will evolve from executing assigned tasks to proactively seeking out work. Once Agents handle routine responsibilities, professionals will need to rethink what non-routine problems they can uniquely solve.