The Autonomy–Intelligence Compass- Mapping the Agentic Finance Landscape in 2026

Decentralised News

Recommended crypto platforms, wallets and trading tools: https://decentralised.news/ecosystem Read the complete guide: https://decentralised.news/mapping-the-agentic-finance-agentfi-landscape-2026  Welcome to the next era of finance — where autonomous AI agents don’t just automate tasks… they think, decide, and execute across DeFi, trading, research, and governance.

Transcript

00:00next evolution of financial technology is here and it's not just about automation it's actually

00:05about autonomy so agentic finance agent fi refers to these financial systems that are powered by ai

00:13agents that can actually manage funds and also execute trades analyze markets and even communicate

00:20directly with users so these aren't simple bots following sort of our preset rules here they are

00:26actually adaptive learning operating within a decentralized financial ecosystem so this year

00:33we're looking at agent fi as an early phase innovative uh kind of a progressive um acceleration

00:41of ai with the crypto space developers traders and investors are asking you know kind of the same key

00:49questions here how smart are these agents and how much freedom do they actually have and to answer

00:54that we will be introducing kind of like this uh idea of our autonomous intelligence uh campus

01:01which is a framework that kind of maps uh the diverse range of age and five projects are there

01:07by kind of defining two traits i mean two defining traits basically the intelligence and the actual

01:13autonomy itself so understanding that autonomy intelligence uh campus is very important let's

01:19look at the kind of the x-axis intelligence which is a measurement of how advanced these agent um

01:26decision making capabilities are right so on the far kind of uh left of uh kind of these rule-based

01:34agents using uh mathematical statistical logic right in the middle you've got uh machine learning

01:40agents that identify patterns and optimize decisions over time and also on the far right you've got these

01:46language uh models uh the llm lms uh agents which are combined reasoning uh communication adaptive

01:55strategy and also kind of how they're capable of acting uh more like human financial analysts so on

02:01the y-axis you kind of have the autonomy side where you know kind of measuring how independently

02:06these agents operate at the bottom of that maybe we're looking at a kind of advisory agents right

02:12that provide insights and they don't actually act without any approval in the middle you've

02:16got the human in the loop agents kind of um which i act uh semi-autonomously require user

02:23confirmation for critical steps but at the top of these fully autonomous agents of course executing

02:28trades and managing capital directly on chain without any human input most agent five products

02:34actually sit today in the middle zone right they kind of relatively intelligent semi-autonomous and

02:39evolving quickly towards the fully autonomous uh kind of confident uh um systems you know so

02:45mapping uh kind of the landscape of this agentic finance is very important because obviously

02:51you've got different categories within that you've got the trading and portfolio agents kind of uh

02:57machine learning hybrid autonomous uh agents highly autonomous uh automated asset allocation is what

03:04they do multi uh kind of a chain execution then you also have liquidity provisioning agents which are

03:10rule-based and uh kind of high autonomy as well optimizing liquidity positions and minimizing uh impermanent

03:18losses and uh you've got late lending agents uh which are also kind of have the rule-based intelligence

03:24here minimum medium autonomy selecting yield strategies managing loan exposure for example and then you've

03:30got prediction betting agents hybrid lms uh medium to high autonomy the sports prediction markets

03:38uh finance um kind of info finance strategies uh basically and then you've got analysis and research

03:44agents also uh they are llmz large language models uh with low medium autonomy sentiment fundamentals alpha

03:52discovery that kind of thing that's kind of what they look at but we look at some of the key players

03:57in their positions on this campus obviously you've got the virtual protocol which is on base a fully uh

04:03uh full full ecosystem of uh ai agents on chain enabling developers to launch a co-owner intelligent

04:11revenue generating agents and these agents actually communicate and they also react transact and even

04:18kind of form micro economies are placing virtuals near the top right kind of quadrant high autonomy high

04:25intelligence aside then you've got the you know projects like banker uh which represent uh practical

04:32consumer facing uh side of the agent five chat based defy assistance you know that execute commands

04:37directly uh from text or social media uh obviously while the the intelligence uh blends rule-based logic

04:44and uh convince uh conversational ai kind of operates a semi-autonomous uh sitting in the middle

04:51zone and then you've got agent fi uh similar platforms also kind of like a self-chain or autonomous

04:58building frameworks that agents interact securely transparently on and then you've got these solana

05:03trading bots for example high-end high-speed algorithmic parts of solana exhibiting high autonomy

05:10but low intelligence powerful executors not really true decision makers and uh they kind of occupy that

05:18uplift quadrant right showing kind of what autonomy alone doesn't equate to intelligence and then you know

05:23most agents i think remain hybrid because obviously you've got the model limitations right uh lms are

05:30prone to errors still uh inconsistencies inconsistencies in their arithmetic and logic and

05:36unacceptable risks from managing real capital as well regulatory incentives as well autonomous financial

05:42systems play accountability right because until these legal frameworks around there might be a need for

05:48some human insight uh oversight uh security and uh custody risks granting full access to you know

05:55really needs for bus safeguards as well so that's uh something that you know it has to be considered

06:00but obviously i think that that landscape is really um expanding in many ways and the trading and

06:05portfolio management space and the liquidity and yield space so in the performing trading portfolio

06:11management you've got uh ai agents like banker wayfinder which automates swaps bridges rebalancing across

06:19chains you've got liquidity yield agents like giza camino optimizing uh lp and stablecoin

06:25strategies autonomously you've got prediction markets obviously like billy bets and polytrader combining

06:30air i forecasting with event-based wagering and you've got data sentiment agents also ai xpt default

06:37t2 using lms to process social on-chain fundamental data for alpha execution detection as well so

06:44together these uh represent kind of this maturing ecosystem where we're seeing these ai agents just

06:50not being not being science fiction anymore but actually evolving into financial entities that are

06:55managing real capital and real assets on chain so i think what's next is really interesting you've got

07:00these hybrid learning models reinforcement learning combined with symbolic reasoning while i kind of

07:06allowing agents to learn from experience as well without a hallucination right agent to agent

07:11markets soon agents will be able to transact negotiate and share liquidity with one another

07:16autonomously verifiable autonomy new protocols logging and auditing agent decisions on chain

07:22ensuring transparency even when no human is directly uh kind of involved in then you've got the

07:27cross-chain uh cognition as well agents will coordinate basically across ethereum based solana matching

07:33execution execution with speed i think three to five years we can expect generation of uh fully self

07:38learning fully autonomous on chain fund managers emerging blending llm uh reasoning on chain risk

07:44control we're very viable uh you know kind of performance so guys if you want to learn more about

07:50ai agents make sure you just tap in centralized don't use is the best place to get all the free tools

07:55resources all the free alpha join all our socials we are available on more than 30 plus socials if you want to

08:01learn more about decentralized finance go get my book d5 millionaire which is a d5-millionaire.com

08:07also check out my other book block uh blockchain applied and my other one tokenized trillions

08:12you can learn about real world asset organization on the blockchain until next time

Category

Transcript

Be the first to comment

Recommended