AI & Crypto: Bots, Tokens & Treasuries
Explore how autonomous agents and tokenized Treasuries are reshaping DeFi and TradFi. Learn what founders and investors need to know
Autonomous agents are here. They’re reallocating portfolios in your sleep, turning chat feeds into wallets, and reshaping how digital value moves. Meanwhile, institutions are quietly tokenizing everything from U.S. treasuries to real estate. This newsletter dives deep into AI × Crypto, a narrative that won’t die because it bridges two megatrends. Markets may wobble, but the machine economy and tokenized finance keep advancing.
Crypto markets look fatigued. AI‑linked tokens slumped 6% this week and opportunistic treasuries are rotating into fringe coins . Yet under the red candles, something bigger is brewing. AI agents now execute trades from within your Telegram chats , and major asset managers are issuing tokenized money‑market funds on public blockchains . The lines between DeFi and TradFi are blurring, and the battle between bots and bankers is just getting started.
Why It Matters
For builders, this shift opens a greenfield: conversational wallets, agent marketplaces, programmable asset wrappers, and trust‑layer infrastructure. For investors, it demands discernment — separating durable platforms from “AI‑in‑name‑only” pumps, and understanding how tokenized yields and autonomous strategies will reshape portfolios. Ignore the narrative at your own risk: AI‑driven DeFi and tokenized treasuries are not competing stories but complementary pillars of the next financial stack.
Deep Dive: AI × Crypto and the New Finance
Market Snapshot: Hype versus Reality
The combined market cap of AI‑related crypto tokens sits in the low single‑digit billions. Hype cycles have pushed and pulled valuations — some names shot up ten‑fold during ChatGPT’s heyday only to retrace sharply. Recent months have seen the sector lag broader crypto, with AI coins falling faster than Bitcoin and Ether . This weakness reflects speculation outrunning adoption, but it doesn’t negate the underlying thesis. Unlike memecoins, AI token projects represent two intertwined trends: autonomous agents and new incentive models.
A parallel development is playing out in public markets. Digital asset treasuries (DATs) — public companies holding crypto on their balance sheets — are shifting from Bitcoin into more exotic coins in search of yield . Analysts warn that venturing into illiquid tokens increases volatility and pressure on equities when markets drop . One DAT even justified its purchase of NEAR as a way to tap the token’s built‑in AI capabilities . The takeaway: speculation around AI permeates corporate treasuries, but risk management has yet to catch up.
Inside the AI Machine Economy
At the heart of the AI × Crypto narrative are autonomous agents — software entities that hold keys, call contracts, and make decisions for users. They transform DeFi from a manual affair into a set‑and‑forget system. Here’s what’s happening today:
Conversational wallets: Bankr’s on‑chain agent turns a chat handle into a wallet login. You send “Buy $100 of ETH” in Telegram or X; the agent executes the swap, bridges if necessary, and records every transaction on‑chain . Gas fees are abstracted, and safety features like refusing scam tokens are built in . This design meets users where they already spend time, hiding private keys while preserving non‑custodial control.
Agent marketplaces: Autonolas’s Pearl platform lets anyone browse, launch and fund agents on a decentralized network . You define a goal (rebalance a staking portfolio, collect subscription fees, propose DAO votes), fund the agent via card or crypto, and it runs autonomously while every action is recorded on‑chain . Valory co‑founder David Minarsch frames it as giving users “digital teammates” they own, not services they rent .
Chat‑native commerce: Within the TON network, mini‑apps like Goodies blend collectibles, brands and payments inside Telegram. Users buy and trade sticker‑pack NFTs from Kung Fu Panda or Pudgy Penguins without leaving the chat . Each sticker is a token; sending one shares both emotion and ownership . This shows how culture and commerce merge when payments are native to messaging.
These examples illustrate the user‑experience unlock agents provide. People can invest, swap, pay, and earn in the same interface where they post memes or schedule meetings. For builders, the challenge is stitching together off‑chain AI logic with on‑chain identity and settlement while maintaining security and compliance.
Infrastructure & Key Players
To deliver on this vision, several layers of infrastructure are emerging:
Agent frameworks: OpenAI recently released Responses API, built‑in tools and an Agents SDK. These building blocks let developers compose single‑ or multi‑agent workflows, leveraging built‑in functions like web search, file search and computer use . For startups, this means less time building orchestration logic and more time focusing on core features.
Agent networks: Autonolas’s Pearl functions like an app store for agents. It ensures each task is verifiable — every action is logged on‑chain, so owners can audit performance . The network’s decentralized design keeps agents running even if parts of the system go offline .
Conversational wallets: Bankr is more than a bot; it’s an agentic payment layer embedded in social feeds. Its integration with Privy handles sign‑ins via Telegram or X, meaning users don’t juggle seed phrases. When you type a command, the agent can bridge assets, place limit orders, settle Polymarket bets and list NFTs . Fail‑safe features reject known scam tokens, building trust .
Tokenized funds & RWA rails: On the institutional side, big players are building infrastructure for tokenized assets. Franklin Templeton’s tokenized money‑market fund (BENJI tokens) now spans four regions — the U.S., Luxembourg, Singapore and Hong Kong — with assets exceeding $900 million . The firm uses a proprietary on‑chain transfer agent to issue and track shares . UBS went further, completing a live, in‑production tokenized fund transaction on Nov. 4. Its uMINT money‑market fund, built on Ethereum, handles subscription and redemption via Chainlink’s Digital Transfer Agent standard . This workflow automates order taking, execution, settlement and data synchronisation across chains .
The common thread across these platforms is programmability. Agents need secure rails to move value; tokenized funds require smart contracts to enforce compliance and handle settlement. As more assets become digital, the infrastructure to support autonomous agents and RWA tokenization converges.
Risks and Realism
While the promise is huge, several risks loom:
Speculative cycles: Many AI tokens underperform after initial hype. When companies and traders chase narratives without underlying adoption, volatility spikes and retail investors get burned. DATs venturing into exotic tokens amplify these swings .
Technical maturity: Current agents are brittle. They rely on off‑chain AI models that can hallucinate or misinterpret instructions. On‑chain smart contracts demand deterministic inputs, so any fuzziness can lead to costly errors. Building robust interfaces between probabilistic AI and deterministic blockchains remains a work in progress.
Security: Agents hold keys and execute financial transactions; this makes them attractive targets. Without rigorous audit and runtime monitoring, a compromised agent could drain funds or act maliciously. Builders must bake in fail‑safes, monitoring and kill switches.
Regulation: Autonomous agents raise complex questions. If a bot trades for thousands of users, is the operator an unregistered broker? Do payments agents need money‑transmitter licences? Tokenized funds might fall under securities laws. Regulators are still catching up. The NEAR example — touting AI capabilities to justify investment — shows how marketing can race ahead of legal clarity .
RWA vs. AI: Twin Rails for Autonomous Finance
Alongside AI agents, real‑world asset (RWA) tokenization has become one of crypto’s hottest themes. On the surface they seem distinct: RWAs bring traditional assets on‑chain, while AI projects build new on‑chain services. In reality, they reinforce each other.
Institutional bridge: Franklin Templeton’s BENJI tokens and UBS’s uMINT fund demonstrate that major asset managers see value in moving Treasuries onto public chains . These products offer 24/7 settlement, transparency and fractional ownership — benefits that align with DeFi’s ethos even if the issuers remain centralized.
Machine uptake: AI agents need reliable, yield‑bearing assets to manage. Tokenized treasuries provide a stable, on‑chain instrument that bots can incorporate into portfolios or use as collateral. When a conversational wallet can park idle funds in a tokenized money‑market fund with one command, DeFi starts to look like a robo‑advisor.
Decentralization trade‑offs: RWAs often operate on permissioned rails to satisfy compliance. Agents like Bankr and Pearl strive for permissionless autonomy. The tension between open systems and controlled assets will shape how products are designed. Builders must decide when to embrace centralized interfaces (for regulatory reasons) and when to push for on‑chain independence.
The bottom line: AI agents automate the how, while RWA tokenization expands the what. They’re not competing narratives but two legs of the same stool.
Actionable Insights
For founders building in this space:
Simplify the UX: Users don’t care about your protocol; they care about outcomes. Conversational wallets and no‑code agent studios lower friction. Focus on meeting users where they already are.
Leverage new frameworks: Tools like OpenAI’s Responses API and Agents SDK reduce the burden of building multi‑agent orchestration . Combine them with on‑chain state machines to create resilient services.
Plan for compliance: Even if your project is fully decentralized, your users are in regulated jurisdictions. Consider partnering with entities that can offer regulated wrappers (e.g. tokenized funds) while keeping the agent logic open source.
For investors and ecosystem VCs:
Look beyond the label: Scrutinize whether a token’s AI narrative is backed by functioning agents or just marketing. Real usage — such as Bankr processing trades or Pearl executing DeFi strategies — matters more than ticker symbols.
Assess risk exposure: DATs and treasuries venturing into volatile tokens highlight the risks of reaching for yield . Diversifying into tokenized treasuries or stablecoins might offer better downside protection while retaining on‑chain composability.
Watch for convergence: The best opportunities may lie at the intersection — funds that enable bots to hold RWAs, or agent marketplaces that monetize data feeds from tokenized assets.
Forward‑Looking Questions
Does sentiment die when bots trade with bots? When AI agents dominate order flow, will markets become hyper‑efficient or experience new kinds of flash crashes? Does human sentiment still matter if most volume is machine‑driven?
What rights should agents have? As bots hold assets and vote in DAOs, should there be disclosure or “know‑your‑machine” rules? Who’s liable when an autonomous fund loses money or manipulates prices?
How far will tokenization go? If treasuries and funds move on‑chain, what about equities, mortgages, or copyrights? Will AI agents one day manage a portfolio of tokenized real estate and art alongside yield‑bearing crypto?
These questions will shape the next phase of crypto and AI. If you’re building or investing, now is the time to engage.
Ecosystem Headlines
AI tokens slump 6% as treasuries pivot to exotic bets. Digital‑asset treasury companies are rotating out of Bitcoin into more volatile tokens, highlighting the risk of chasing yield. One firm even justified its NEAR purchase by touting the token’s AI integration . The broader AI‑token sector fell roughly 6% during the week, underscoring how narrative‑driven markets can reverse quickly.
Franklin Templeton’s tokenized funds top $900M and reach Hong Kong. The asset manager expanded its OnChain U.S. Government Money Fund to Hong Kong, the fourth region after the U.S., Luxembourg and Singapore. BENJI tokens now exceed $900 million in market capitalization, proving institutional demand for on‑chain money funds .
🔗 https://www.ledgerinsights.com/franklin-templeton-launches-tokenized-mmf-in-hong-kong/
UBS completes first live tokenized fund transaction on Ethereum. On Nov. 4 the Swiss bank processed subscriptions and redemptions for its uMINT money‑market fund via Chainlink’s Digital Transfer Agent. The workflow automated order taking, execution and settlement across chains . UBS manages $6.9 trillion in assets; this move signals that mainstream finance is embracing on‑chain settlement .
🔗 https://www.coinspeaker.com/swiss-bank-first-tokenized-fund-ethereum/
Chat‑native wallets and agent stores gain traction. Bankr’s conversational wallet lets users buy crypto, bridge assets and even settle wagers directly from X or Telegram . Autonolas’s Pearl agent marketplace enables anyone to launch a custom agent that automates repetitive financial tasks . Together they show a future where UX, not just protocol design, wins adoption.
Builder’s Corner
Bankr – Conversational DeFi Wallet. Turn your X or Telegram handle into a non‑custodial wallet. Type “Buy $100 of ETH” and Bankr’s agent executes the trade, bridges chains and records it on‑chain . Safety checks refuse scam tokens . Great for builders exploring chat‑native commerce.
🔗 https://bankr.bot
OpenAI Responses API & Agents SDK. OpenAI’s new API primitive combines chat completions with tool use, plus built‑in web search, file search and computer‑use tools. The Agents SDK orchestrates multi‑agent workflows and adds observability . It’s perfect for developers building on‑chain agents that need off‑chain reasoning.
Autonolas Pearl – Agent App Store. Pearl lets anyone spin up an AI agent to manage staking, execute trades or handle payments. Agents are fully owned by the user, with all actions logged on‑chain . Ideal for teams who want turnkey autonomy without surrendering control.
🔗 https://olas.network
📢 Cracked Labs Spotlight
Applications closing Nov 23rd. Somnia’s Dreamathon is coming to a close. Is your favorite project going to make it to demo day?
Demo Day Dec 5th. Our current cohort will showcase prototypes ranging from DeFi trading agents to on‑chain supply‑chain trackers. Reserve your (virtual or in‑person) seat now; investors from top funds will be attending.
Alpha Leak
Most of the team will be missing DevConnect this year, sad we know, however we do have some people on the ground. If you find them and take a picture we will send you a Cracked Labs t-shirt. Go ahead, ask around.
Questions? Want your project featured? Reply to this email or ping us on Telegram.