Cyber Agent launches AI contest with ¥10M prize, but rejects feasibility demands

Cyber Agent, a major Japanese digital marketing and media company, is undertaking a comprehensive effort to embed AI thinking across its workforce. The impetus dates to 2016, when the company began exploring AI applications in earnest. However, the launch of ChatGPT in 2022 created a turning point: the technology's accessibility and raw capability sparked renewed urgency to scale AI adoption not just within specialist teams but across all business units and job functions.

To that end, the company is running a second iteration of an "AI Use-Case Contest" beginning in February 2025. The contest carries a ¥10 million prize pool for winning ideas. Crucially, the rules state that "feasibility is not required." This departure from typical corporate innovation contests signals Cyber Agent's belief that employees often self-censor ideas they fear are impractical or outside their scope. By removing feasibility as a gate, the company aims to surface raw creativity about how AI could reshape work. The contest solicits ideas from all divisions and employee levels, not just technologists. Winning ideas then proceed through multiple refinement phases. Employees who propose ideas are expected to design not just the concept but also the implementation process—how the idea would be introduced, how it would create value for the business and for the broader organization, and how it would eventually roll out into live practice. The first results are published in October 2025, with final certification by April 2026. Thereafter, Cyber Agent intends to run the cycle annually.

Parallel to the contest, Cyber Agent has launched an "AI Practice Certification" program that ranks employees across a defined capability ladder. The framework assesses how deeply an employee understands AI and how effectively they apply it within their role, team, and organization. Levels span from "beginner" (recognizing the value of AI at an individual level) through "intermediate" (applying AI within one's team), "leader" (driving organizational adoption), and "innovator" (pioneering new use cases and thinking). The company publishes rankings in October 2025 and repeats annually. This visible ranking system serves multiple purposes: it signals that AI capability is now a valued and measurable competency; it creates social and career incentives for employees to invest in AI literacy; and it gives leaders a transparent view of how AI adoption is distributed across the organization.

Freee, a financial software company also featured in Cyber Agent's AI initiatives, offers a parallel example of this adoption model. Freee has created an "easy-to-use AI bot" to lower the barrier to entry for non-technical employees. The bot surfaces documentation from Confluence and integrates with Slack, allowing employees to ask questions in their native workflow. Employee feedback is fed back into the bot's learning loop, creating a cycle of continuous improvement. Freee's Chief Innovation Officer notes that the bot is treated as a tool that learns alongside users, not as a finished product—a mindset that reflects the broader organizational shift toward treating AI as an evolving capability, not a one-time implementation.

Cyber Agent's leadership has framed the strategy around a single principle: "AI is not something to apply blindly, but a capability we must understand and use wisely." This reflects recognition that AI's competitive value lies not just in deploying the latest models, but in how deeply an organization embeds AI reasoning into decision-making, role design, and culture. By running contests that reward creative thinking and certifications that measure organizational capability, Cyber Agent is signaling that the next phase of AI adoption is about human judgment and organizational readiness, not just technology.