How Generative AI is Transforming Hiring Strategies and Business Norms

The “Qualitative Shift in Hiring” Behind the Starting Salary Bubble

Starting salaries for new graduates are rising across the board. Major IT companies are increasingly offering monthly salaries exceeding 300,000 yen (approx. $2,000), creating a situation that can only be called a “starting salary bubble.”

However, the essence of this news is not the surge in salary levels. According to ITmedia reports, one major company has adopted a strategy of reducing the number of hires while using AI to cover tasks, based on the judgment that “generative AI is more capable than new graduates.”

This is not merely cost-cutting. It represents a paradigm shift in management, redefining the balance between “talent quality” and “AI utilization.”

Speaking from my own experience of combining multiple AIs like Claude, ChatGPT, and Grok in my company to achieve a reduction of 1,550 work hours annually, I can say this trend is not limited to a few large corporations.

Three Management Messages Behind “Ultra-Selective Hiring”

Behind ultra-selective hiring lie the following three management decisions.

First is “optimizing personnel costs.” If a monthly AI tool costing a few tens of thousands of yen (a few hundred USD) can produce output equal to or better than hiring ten new graduates at 300,000 yen ($2,000) each, the return on investment is clear.

Second is “redesigning business processes.” The idea of clearly separating routine tasks that can be handled by AI from creative tasks that only humans can perform is gaining traction.

Third is “organizational streamlining.” A structure where a small, elite team of employees leverages AI to manage operations directly contributes to reducing indirect costs.

What Pasona’s 9,000+ AI Users Reveal About “Internal Adoption Reality”

Alongside changes in hiring strategy, the level of AI adoption within companies is also noteworthy. At Pasona Group, internal AI use has advanced to the point where their generative AI community has been recognized with awards. Over 9,000 employees are using AI in their work, achieving concrete improvements.

The key takeaway from this case is that the success of AI implementation depends not on “tool performance” but on “organizational culture.” Rather than forcing AI adoption top-down, Pasona fostered a community where employees voluntarily learn from each other.

Three Steps for AI Adoption Achievable Even by SMEs

The three steps I recommend to client companies are as follows.

First, “accumulate small successes.” Before aiming for company-wide implementation, conduct a trial run of AI-based process improvements in a specific department. Start with tasks where the effect is easily visible, such as invoice processing in accounting or email drafting support in sales.

Second, “share and horizontally deploy success stories.” Pasona’s community-based approach is highly effective in this regard. Create an AI utilization channel on Slack or Teams within your own company to provide a space for employees to share know-how.

Third, “implement AI literacy training.” Regularly hold sessions not just on how to use the tools, but also on skills like critically evaluating AI output and the basics of prompt design.

New Company for Domestic AI Development and the “Information Accuracy” Challenge

A new company for domestic AI development, backed by over 15 firms including Toshiba and Hitachi, has been established. This move signals that companies are beginning to view AI not as an “external service” but as a “core part of their management infrastructure.”

At the same time, the challenge of “information accuracy” in using generative AI has come into sharp focus. While one survey shows 86.7% of users feel the effects are real, concerns about hallucinations (generation of false information) remain deeply rooted.

Practical Approaches to Ensuring Information Accuracy

Based on my experience, rather than blindly trusting AI output, you can ensure practical accuracy by establishing the following three checkpoints.

First, “systematize fact-checking.” When using AI for critical tasks, always incorporate a process for final human review. Especially for contract reviews and legal matters, AI suggestions should not be taken at face value; they must be checked by an expert.

Second, “develop the habit of comparing AI outputs.” Don’t rely on just one AI. Ask the same question to both Claude and ChatGPT and compare the answers to identify biases.

Third, “continuously improve prompts.” Even for the same question, the way you write the prompt can significantly change the accuracy of the output. Build a system to accumulate and share prompts optimized for your company’s specific tasks.

AI Agents Are Already in Practical Use: What Executives Should Do Now

According to Nikkei BP reports, the corporate adoption of AI agents has already entered a widespread phase. AI agents that autonomously execute tasks, rather than being simple chatbots, are further accelerating business automation.

In my own company, I operate 32 AI agents that automate a wide range of tasks, from social media posts to contract reviews and FX trading. The monthly cost is about 21,000 yen ($140), but they generate value equivalent to approximately 7.53 million yen ($50,000) annually.

Three Actions to Start Right Now

As a business leader, here are three actions you should take immediately.

First, “take stock of which tasks can be replaced by AI.” List all your business processes and classify them into routine tasks that AI can handle and creative tasks that only humans can perform.

Second, “estimate the ROI of AI adoption.” Calculate the specific numbers: how many employees’ worth of work can a monthly AI tool costing a few tens of thousands of yen (a few hundred USD) replace?

Third, “start small and verify the effects.” Don’t rush into company-wide implementation. Start with a specific department or task, measure the results, and then expand.

Conclusion: AI Utilization IS Your Talent Strategy

The judgment that generative AI is more capable than new graduates is by no means an extreme opinion. In reality, with proper AI utilization, it is entirely possible for one employee to handle the work of three or even five people.

However, this does not mean “humans are unnecessary.” On the contrary, the value of people who can effectively leverage AI is higher than ever. What ultra-selective hiring signals is a shift towards “quality” over “quantity.”

What is required of executives is to position AI not merely as a tool for operational efficiency, but as the core of their management strategy. In hiring, training, and business design, the time to start building an organization that assumes AI integration is now.