With 1.39 billion dollars raised and over 17,000 AI agents launched, AI has become one of the dominant narratives of the decentralized ecosystem.
Since the beginning of the year, on-chain activity related to AI has grown by 86%, reaching about 4.5 million unique active wallets every day. This figure brings AI’s market share to 19%, just behind gaming (20%), marking a significant jump from 9% in January.
It’s not just about fleeting enthusiasm: this growth reflects a structural change in the way users interact with decentralized applications.
AI agents are emerging as a new layer of interaction, capable of automating, optimizing, and acting on behalf of users, without replacing them but extending their possibilities.
What are AI Agents?
AI agents are autonomous software programs that perform tasks, make decisions, or interact with users based on goals, prompts, or real-time data.
If in the traditional world these agents find application in sectors such as finance, healthcare, and customer support, in Web3, “crypto-native” agents with specific roles and capabilities are emerging.
In detail, some AI agents operate as copilots DeFi, managing yield strategies or executing trades; others are social agents representing users in decentralized social networks, managing profiles and interactions.
In gaming, native game agents are emerging, trained on lore, mechanics, and player behaviors, becoming intelligent guides or opponents.
This evolution is not only theoretical: according to the data from cookie.fun, 1,748 AI agents are already active in different environments.
Only on Virtuals Protocol, a platform that allows the creation and deployment of AI agents, over 17,000 agents have been launched since November 2024, with an average of more than 85 new agents per day.
The real number is probably higher, given that more and more blockchains are investing in infrastructures for the creation, training, and deployment of these agents.
Tokens: utility, hype e capitali
The majority of AI agents in Web3 is accompanied by a token.
These tokens can have multiple functions: powering the governance of the community, enabling advanced features, or funding development, since training and maintaining AI agents requires significant resources.
In other cases, the token is primarily a capitalization tool, useful for creating liquidity, rewarding early users, or riding the market wave. There are also tokens more similar to memecoin, launched on weak narratives but capable of attracting speculation.
Despite the volatility, the market for tokens linked to AI agents shows significant numbers: the total capitalization is 5.9 billion dollars, with 1.4 billion dollars traded in the last 24 hours.
However, the sector has undergone a strong correction: at the beginning of June, the capitalization was 16.6 billion, with a decrease of 64% in a few weeks, reflecting the general market conditions and the typical volatility of emerging sectors.
The AI agents are here to stay, but the relationship between real utility and hype of the tokens still needs to mature.
The success of AI agents largely depends on the underlying infrastructure. The blockchains that support high-volume AI dapps are laying the groundwork for the spread of autonomous agents.
Between January and June 2025, the most active chains for the use of AI dapps were:
– Matchain, with almost 1.9 million daily users, driven by social and agent-centric dapps.
– opBNB and Nebula, which show strong usage thanks to lightweight and gamified AI services.
Even if not all these dapps already employ autonomous agents, the trend is clear: as the frameworks for AI agents mature, these chains will be ready to welcome the next wave of autonomous agents, whether in the DeFi, gaming, or social sectors.
A global revolution: where do the users come from?
The AI agents live on-chain, but their users are distributed all over the world. Analyzing the origin of the users helps to understand adoption trends and market opportunities.
In the first half of 2025, Europe leads the interaction with AI dapps, representing 26.2% of sessions. It is followed by Asia with 21.9% and North America with 15.8%.
South America is still marginal (2.5%), but growing. A significant 33% of interactions come from unspecified or hard-to-categorize regions, including users who use VPNs or come from less-tracked geographical areas.
This global spread demonstrates that AI agents are not a localized phenomenon: the demand is cross-cutting and rapidly expanding across all continents.
The narrative of AI dominates not only conversations but also investment flows. In 2025, projects related to AI agents raised 1.39 billion dollars, with an increase of 9.4% compared to 2024.
This data indicates the growing confidence of investors in the potential of autonomous on-chain agents, which are establishing themselves as one of the most promising verticals of Web3, now capable of rivaling historically strong sectors like blockchain gaming.
Although the figures are still far from investments in centralized AI, the trend is clear: investors see AI agents as a new foundational infrastructure, destined to redefine interaction with decentralized protocols.
The rise of AI agents marks a profound change in the way we interact with decentralized systems. From DeFi copilots to gaming companions, these agents are rapidly transitioning from experimental status to fundamental infrastructure.
The numbers are clear: over 17,000 agents launched on Virtuals Protocol, a token market worth 5.9 billion dollars and 1.39 billion raised just this year. Adoption is global, with Europe, Asia, and North America at the forefront.
Challenges remain to be faced: many tokens are still driven more by hype than by utility, and not all agents keep their promises of autonomy.
However, as tools mature and concrete use cases emerge, the agent economy is increasingly approaching a new normal, where interaction with on-chain AI will be the rule, not the exception.
The race has just begun: building smarter agents, stronger ecosystems, and clearer standards will be the key to the future of Web3.
