The proportion of AI in overall startup investment is expected to double by the end of the year, up from 10% three years ago . This year, it has expanded into vertical areas in addition to horizontal areas. As the resolution of industries has increased, more AI can be used in business .
AI agents that autonomously run workflows can control output quality, and as a result, they have been rapidly gaining popularity in enterprise applications since 2025. As a result of this, progress is currently being made in developing orchestration systems that bundle multiple AI agents together, guardrails to prevent compliance violations, and security frameworks to protect confidential information.
Robotic speech is a thing of the past, and current voice AI has evolved to a level where it is difficult for the average person to distinguish, and it is predicted that voice AI will play a leading role from the second half of 2025. While voice communication is ideal, a trend toward a return to voice is beginning to emerge in fields that are text-based due to labor shortages (such as customer support and outbound sales).
The multi-agent system recreates the individual personalities of each individual human being in an AI agent, allowing them to interact with each other just like humans would, simulating what the future might hold. It is expected to be used in a wide range of fields, from predicting sales of new products to voting behavior in elections and evacuation guidance in the event of a disaster.
LLMs (large-scale language models) have given way to SLMs (small-scale language models), and now a new challenge called LWMs (large-scale world models) is emerging. The ultimate goal is a worldview of spatial intelligence that combines intelligence with AI -generated spaces , but the size of the market is still unknown.
Report format: PDF (8.9MB)
Original data: PowerPoint, 80 slides, A4 size