{"id":6343,"date":"2026-02-12T10:27:40","date_gmt":"2026-02-12T10:27:40","guid":{"rendered":"https:\/\/cryptolinks.com\/news\/?p=6343"},"modified":"2026-02-12T10:27:40","modified_gmt":"2026-02-12T10:27:40","slug":"ai-ethereum-is-getting-real-fast","status":"publish","type":"post","link":"https:\/\/cryptolinks.com\/news\/ai-ethereum-is-getting-real-fast","title":{"rendered":"AI + Ethereum Is Getting Real Fast: Vitalik\u2019s ETH Vision and What It Means for DeFi Over the Next 6 Months"},"content":{"rendered":"<p><strong>What happens when an AI agent can actually use DeFi<\/strong>\u2014swap, borrow, hedge, rebalance, and even pay for compute\u2014<em>without you hovering over every click like an anxious co-pilot<\/em>?<\/p>\n<p>Because that\u2019s the shift I\u2019m watching right now. Not \u201cAI writes market summaries\u201d (yawn). I mean <strong>AI that can execute<\/strong>\u2014onchain, with rules, logs, and accountability\u2014so it can operate like a real economic actor.<\/p>\n<p>And if that execution layer ends up being <a href=\"https:\/\/cryptolinks.com\/layer-2-blockchains\">Ethereum (plus L2s)<\/a>, Vitalik\u2019s recent \u201c<a href=\"https:\/\/cryptolinks.com\/crpyto-ai\">ETH powering AI<\/a>\u201d angle isn\u2019t just a narrative. It\u2019s a roadmap for <strong>new onchain activity, new fee flows, and new DeFi products<\/strong> that are hard to build in the old \u201chuman clicking buttons\u201d world.<\/p>\n<p><strong><em>Listen to this article:<\/em><\/strong><\/p>\n<audio class=\"wp-audio-shortcode\" id=\"audio-6343-1\" preload=\"none\" style=\"width: 100%;\" controls=\"controls\"><source type=\"audio\/mpeg\" src=\"https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/Article-read-AI-Ethereum-Is-Getting-Real-Fast-Vitaliks-ETH-Vision-and-What-It-Means-for-DeFi-Over-the-Next-6-Months.mp3?_=1\" \/><a href=\"https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/Article-read-AI-Ethereum-Is-Getting-Real-Fast-Vitaliks-ETH-Vision-and-What-It-Means-for-DeFi-Over-the-Next-6-Months.mp3\">https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/Article-read-AI-Ethereum-Is-Getting-Real-Fast-Vitaliks-ETH-Vision-and-What-It-Means-for-DeFi-Over-the-Next-6-Months.mp3<\/a><\/audio>\n<blockquote><p><em>AI is great at deciding. Crypto is great at executing and proving what happened.<\/em> The fusion is happening because each side fixes the other\u2019s biggest weakness.<\/p><\/blockquote>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-6353\" src=\"https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/The-real-pain-points-why-DeFi-and-AI-both-hit-walls-without-each-other.png\" alt=\"The real pain points why DeFi and AI both hit walls without each other\" width=\"1536\" height=\"1024\" srcset=\"https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/The-real-pain-points-why-DeFi-and-AI-both-hit-walls-without-each-other.png 1536w, https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/The-real-pain-points-why-DeFi-and-AI-both-hit-walls-without-each-other-300x200.png 300w, https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/The-real-pain-points-why-DeFi-and-AI-both-hit-walls-without-each-other-1024x683.png 1024w, https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/The-real-pain-points-why-DeFi-and-AI-both-hit-walls-without-each-other-768x512.png 768w\" sizes=\"auto, (max-width: 1536px) 100vw, 1536px\" \/><\/p>\n<h2>The real pain points: why DeFi and AI both hit walls without each other<\/h2>\n<p>DeFi already has the raw ingredients: liquidity, composability, and 24\/7 global access. But it\u2019s still too manual, too fragmented, and too risky for normal people.<\/p>\n<p>AI has automation and decision-making. But AI, by default, is a black box that can be tricked\u2014and it doesn\u2019t come with native rails for <strong>permissioned action, transparent logs, or economic accountability<\/strong>.<\/p>\n<p>Put them together the right way, and you get a system where:<\/p>\n<ul>\n<li><strong>AI reduces friction<\/strong> (less \u201cpro mode\u201d clicking and babysitting).<\/li>\n<li><strong>Ethereum reduces trust requirements<\/strong> (verifiable execution, audit trails, and composability).<\/li>\n<\/ul>\n<p>The key phrase is \u201cthe right way,\u201d because automation also increases the blast radius when things go wrong. I\u2019ll get to that.<\/p>\n<h3>DeFi is still \u201cpro mode\u201d (and normal users know it)<\/h3>\n<p>If you\u2019ve been in DeFi for a while, you forget how insane the default workflow looks to a normal person:<\/p>\n<ul>\n<li>Pick a wallet, secure seed phrases, avoid fake extensions.<\/li>\n<li>Bridge to the \u201cright\u201d chain\/L2 and pray the route is safe.<\/li>\n<li>Understand collateral ratios, liquidation thresholds, and interest models.<\/li>\n<li>Sign transactions that read like legal disclaimers written by robots.<\/li>\n<li>Track rates that change hourly and positions that can explode overnight.<\/li>\n<\/ul>\n<p>That\u2019s why DeFi adoption keeps hitting the same ceiling: even people who <em>want<\/em> the yields and the autonomy don\u2019t want the constant attention tax.<\/p>\n<p>And it\u2019s not just vibes\u2014user studies repeatedly show that crypto UX complexity is a core blocker. For example, the Ethereum Foundation\u2019s user research work (a good starting point is the EF research hub at ethereum.foundation\/research) and multiple wallet usability studies point to the same pattern: people struggle with transaction comprehension, risk visibility, and irreversible actions.<\/p>\n<p><strong>This is exactly where agents shine.<\/strong> Not as \u201cgenius traders,\u201d but as tireless operators that can:<\/p>\n<ul>\n<li>Monitor rates and collateral 24\/7.<\/li>\n<li>Rebalance when predefined rules trigger.<\/li>\n<li>Route swaps intelligently across venues.<\/li>\n<li>Batch actions to reduce mistakes and fees.<\/li>\n<\/ul>\n<p>In plain terms: DeFi needs a \u201cdriver assistance system.\u201d AI can be that\u2014<em>if<\/em> it can execute safely.<\/p>\n<h3>AI agents need verifiable action, not just smart text<\/h3>\n<p>Most \u201cAI in crypto\u201d demos today are glorified chat UIs. They can <em>suggest<\/em> a trade, but they can\u2019t <strong>act<\/strong> in a way you can verify, reproduce, and hold accountable.<\/p>\n<p>That gap matters. The moment an agent touches money, you need things normal chatbots don\u2019t provide:<\/p>\n<ul>\n<li><strong>Identity &amp; permissions<\/strong>: who\/what is this agent allowed to do?<\/li>\n<li><strong>Transaction execution<\/strong>: can it submit valid actions without begging a human every time?<\/li>\n<li><strong>Audit trails<\/strong>: what did it do, when, and why?<\/li>\n<li><strong>Economic accountability<\/strong>: if it misbehaves, can it be penalized (bonding\/slashing\/reputation)?<\/li>\n<\/ul>\n<p>This is where blockchains are just\u2026 built different. Onchain systems naturally produce an append-only history, and smart contracts can enforce rules the agent can\u2019t \u201csweet-talk\u201d its way around.<\/p>\n<p>Also, prompt injection is not theoretical. Tool-using agents can be manipulated through malicious inputs (webpages, documents, even \u201chelpful\u201d instructions). Microsoft has written about real-world prompt injection and data exfil risks for LLM agents here: <a href=\"https:\/\/www.microsoft.com\/en-us\/security\/blog\/\" target=\"_blank\" rel=\"noopener\">Microsoft Security Blog<\/a>. Anthropic has also documented jailbreak\/tool-risk dynamics in their safety research: <a href=\"https:\/\/www.anthropic.com\/research\" target=\"_blank\" rel=\"noopener\">anthropic.com\/research<\/a>.<\/p>\n<p>So if the agent can sign transactions, we need an execution environment that assumes the agent can be tricked\u2014and still keeps the user safe.<\/p>\n<h3>The trust problem: automation increases speed\u2026 and the blast radius<\/h3>\n<p>Here\u2019s the uncomfortable truth: <strong>automation doesn\u2019t just remove friction\u2014it removes friction for failures too.<\/strong><\/p>\n<p>A human trader might make one bad click. An autonomous agent can make 50 bad transactions in a minute if the guardrails are weak.<\/p>\n<p>Real examples of \u201csmall mistake \u2192 big loss\u201d are everywhere in crypto:<\/p>\n<ul>\n<li>Signing the wrong approval (unlimited token approvals are still a common drain vector).<\/li>\n<li>Using the wrong router\/contract address (phishing contracts love speed and distraction).<\/li>\n<li>Following manipulated price feeds or thin-liquidity traps.<\/li>\n<\/ul>\n<p>So the next wave can\u2019t just be \u201cAI executes DeFi.\u201d It has to be \u201cAI executes DeFi <strong>inside a sandbox<\/strong>.\u201d The guardrails I\u2019m watching for in serious implementations look like this:<\/p>\n<ul>\n<li><strong>Spending limits<\/strong> (hard caps per day\/asset\/protocol).<\/li>\n<li><strong>Simulation before execution<\/strong> (preview outcomes, revert if slippage\/price impact spikes).<\/li>\n<li><strong>Intent-based actions<\/strong> (declare <em>what<\/em> you want; let constrained solvers compete on <em>how<\/em>).<\/li>\n<li><strong>Policy engines<\/strong> (rules like \u201cnever interact with unaudited contracts,\u201d \u201cno new approvals,\u201d \u201conly these pools\u201d).<\/li>\n<li><strong>Multisig-style constraints<\/strong> for high-risk moves (agent proposes, human or second key confirms).<\/li>\n<li><strong>Monitoring &amp; kill switches<\/strong> (pause on anomalies, revoke session permissions instantly).<\/li>\n<\/ul>\n<p>If you\u2019ve ever thought, \u201cI\u2019d use DeFi more if it wasn\u2019t so easy to screw up,\u201d you\u2019re not alone. Automation should make DeFi feel safer, not scarier. But that only happens if guardrails are treated as a product feature, not an afterthought.<\/p>\n<h3>Promise solution: what I\u2019ll help you understand in this post<\/h3>\n<p>Here\u2019s what I\u2019m going to do for you in this write-up:<\/p>\n<ul>\n<li>Explain <strong>what people really mean<\/strong> when they say \u201cEthereum will power AI agents\u201d (and what they don\u2019t mean).<\/li>\n<li>Map the practical building blocks that make \u201cAI does DeFi\u201d realistic: wallets, permissions, intent flows, data, execution, monitoring.<\/li>\n<li>Call out the DeFi product categories that are most likely to see real traction soon\u2014<em>not sci-fi<\/em>, the stuff that can ship.<\/li>\n<li>Share a simple checklist I use to spot projects with substance vs projects with a chatbot wrapper and a token ticker.<\/li>\n<\/ul>\n<p>But first, there\u2019s one question you need answered before any of that matters:<\/p>\n<p><strong>When Vitalik says \u201cETH powering AI,\u201d what does that actually look like mechanically\u2014if AI compute isn\u2019t running on Ethereum?<\/strong><\/p>\n<p>That\u2019s where things get interesting, because the answer changes how you evaluate almost every \u201c<a href=\"https:\/\/cryptolinks.com\/\">AI x crypto<\/a>\u201d claim you\u2019ll see this year.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-6352\" src=\"https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/Vitaliks-ETH-powering-AI-idea-\u2014-what-it-actually-means-in-practice.png\" alt=\"Vitalik\u2019s \u201cETH powering AI\u201d idea \u2014 what it actually means in practice\" width=\"1536\" height=\"1024\" srcset=\"https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/Vitaliks-ETH-powering-AI-idea-\u2014-what-it-actually-means-in-practice.png 1536w, https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/Vitaliks-ETH-powering-AI-idea-\u2014-what-it-actually-means-in-practice-300x200.png 300w, https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/Vitaliks-ETH-powering-AI-idea-\u2014-what-it-actually-means-in-practice-1024x683.png 1024w, https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/Vitaliks-ETH-powering-AI-idea-\u2014-what-it-actually-means-in-practice-768x512.png 768w\" sizes=\"auto, (max-width: 1536px) 100vw, 1536px\" \/><\/p>\n<h2>Vitalik\u2019s \u201cETH powering AI\u201d idea \u2014 what it actually means in practice<\/h2>\n<p>When people hear \u201cAI + Ethereum,\u201d they picture robots running inside the EVM like some sci\u2011fi smart contract brain. That\u2019s not the point. The point (and the reason Vitalik\u2019s framing hit so hard) is that Ethereum can be the <strong>neutral coordination and settlement layer<\/strong> for autonomous agents that live offchain, think offchain, but <em>act<\/em> onchain.<\/p>\n<p>Here\u2019s the simplest translation:<\/p>\n<ul>\n<li><strong>AI agents<\/strong> decide what to do (swap, hedge, rebalance, borrow, repay, provide liquidity).<\/li>\n<li><strong>Ethereum + L2s<\/strong> make sure the \u201cdoing\u201d is verifiable, constrained, auditable, and economically accountable.<\/li>\n<li><strong>Smart contracts<\/strong> become the rules, the escrow, and sometimes the judge (disputes, slashing, insurance).<\/li>\n<\/ul>\n<p>If you want the spark that lit up a lot of this conversation, start with Vitalik\u2019s own post and then scan how builders interpreted it:<\/p>\n<p><a href=\"https:\/\/x.com\/VitalikButerin\/status\/2020963864175657102\" target=\"_blank\" rel=\"noopener\">Vitalik\u2019s thread<\/a> \u2022 <a href=\"https:\/\/x.com\/lucianlampdefi\/status\/2021202754329584078\" target=\"_blank\" rel=\"noopener\">Lucian<\/a> \u2022 <a href=\"https:\/\/x.com\/javliscom\/status\/2021412008256491944\" target=\"_blank\" rel=\"noopener\">Javlis<\/a> \u2022 <a href=\"https:\/\/x.com\/0xshai\/status\/2021147938236531180\" target=\"_blank\" rel=\"noopener\">0xshai<\/a><\/p>\n<p>Now let me make this real, with the mechanisms that actually matter.<\/p>\n<h3>Ethereum as the coordination layer for AI agents (not \u201cAI runs on ETH\u201d)<\/h3>\n<p>Most AI compute will not run on Ethereum. It\u2019ll run on GPUs in data centers, on decentralized compute networks, or on your own hardware. What Ethereum (and L2s) can do is handle the parts that AI is terrible at doing safely on its own:<\/p>\n<ul>\n<li><strong>Payments<\/strong>: an agent pays for data, execution, or compute with clean settlement and receipts.<\/li>\n<li><strong>Permissions<\/strong>: who can do what, with which keys, with what limits.<\/li>\n<li><strong>Bonding \/ staking<\/strong>: if an agent or solver misbehaves, it can lose a bond (real consequences).<\/li>\n<li><strong>Dispute resolution<\/strong>: not perfect, but smart contracts can enforce rules and outcomes.<\/li>\n<li><strong>Audit trails<\/strong>: a tamper-resistant log of \u201cwhat happened,\u201d when, and with which constraints.<\/li>\n<\/ul>\n<p>This isn\u2019t theoretical. The DeFi world already learned (the hard way) that <strong>execution environment matters<\/strong>. MEV research like <em>Flash Boys 2.0<\/em> (Daian et al., 2019) made it painfully clear that if you broadcast intent without protection, someone can rearrange the order and extract value. That problem gets worse when execution becomes automated and predictable.<\/p>\n<p>So when Vitalik talks about \u201cETH powering AI,\u201d I read it like this:<\/p>\n<blockquote><p><strong>Ethereum is the court + ledger + payment rail for autonomous economic actors.<\/strong> The \u201cintelligence\u201d lives elsewhere, but the accountability lives here.<\/p><\/blockquote>\n<p>And yes, I\u2019m also watching how the community narrates it across threads like these:<\/p>\n<p><a href=\"https:\/\/x.com\/ThCryptoCook\/status\/2021017173137949138\" target=\"_blank\" rel=\"noopener\">ThCryptoCook<\/a> \u2022 <a href=\"https:\/\/x.com\/Tabl4me\/status\/2021230876563013795\" target=\"_blank\" rel=\"noopener\">Tabl4me<\/a> \u2022 <a href=\"https:\/\/x.com\/houseofai_swan\/status\/2021276364062757069\" target=\"_blank\" rel=\"noopener\">houseofai_swan<\/a> \u2022 <a href=\"https:\/\/x.com\/BSCNews\/status\/2021073501130719374\" target=\"_blank\" rel=\"noopener\">BSCNews<\/a><\/p>\n<h3>The agent stack: what has to exist for \u201cAI does DeFi\u201d to work<\/h3>\n<p>For an AI agent to do DeFi without turning your wallet into a crater, a full stack has to be in place. If any layer is weak, you get \u201ccool demo\u201d energy\u2026 right up until the first ugly loss.<\/p>\n<ul>\n<li><strong>1) Agent logic (models + tools)<\/strong>This is the \u201cbrain\u201d and the toolbelt: route discovery, risk checks, portfolio rules, and a set of approved actions it can call. The key detail: the agent shouldn\u2019t be free-styling transactions. It should be choosing from <strong>bounded actions<\/strong>.<\/li>\n<li><strong>2) Wallet layer (smart accounts, session keys, spending caps)<\/strong>This is where the shift gets serious. With smart accounts (account abstraction), you can give an agent a <strong>session key<\/strong> that can do <em>specific<\/em> things:\n<ul>\n<li>Spend up to X per day<\/li>\n<li>Only interact with whitelisted contracts<\/li>\n<li>Only swap stablecoins for ETH (or the reverse)<\/li>\n<li>Require 2-of-2 confirmation above a threshold<\/li>\n<li>Auto-revoke permissions if conditions trigger (price drop, oracle discrepancy, unusual slippage)<\/li>\n<\/ul>\n<p>This is the difference between \u201cautomation\u201d and \u201chanding your keys to a stranger.\u201d<\/li>\n<li><strong>3) Intent layer (tell the chain what you want, not how)<\/strong>Intents are a big deal for agents because they reduce the number of fragile micro-decisions. Instead of \u201cswap on pool A then bridge then swap again,\u201d the user (or agent) says:<br \/>\n<blockquote><p>\u201cI want to end up with 5 ETH, spending no more than $X, with max slippage Y%, within Z minutes.\u201d<\/p><\/blockquote>\n<p>Then solvers compete to fulfill it. That competition can improve execution, and it can reduce MEV exposure if designed well (this is where the \u201cwho captures the edge?\u201d question gets spicy).<\/li>\n<li><strong>4) Data\/oracles (market data, risk data, identity\/reputation signals)<\/strong>Agents are only as good as their inputs. If the oracle gets manipulated, the agent can \u201crationally\u201d do something irrational. DeFi has been here before: thin liquidity + manipulable pricing has led to painful exploits across multiple cycles.The right direction is <strong>redundant data sources<\/strong>, sanity checks (cross-oracle comparisons), and refusing to act when the data looks wrong.<\/li>\n<li><strong>5) Execution (DEXs, lending, perps, bridges)<\/strong>This is the battlefield. Real liquidity, real slippage, real MEV, real liquidation engines. Agents will use the same venues humans use, but faster and more often\u2014which changes market microstructure.<\/li>\n<li><strong>6) Monitoring + rollback ideas (alerts, circuit breakers, safe-mode)<\/strong>You can\u2019t \u201cundo\u201d blockchain actions, but you can design systems that:\n<ul>\n<li>simulate before sending<\/li>\n<li>pause on abnormal conditions<\/li>\n<li>rotate\/revoke session keys instantly<\/li>\n<li>route into \u201csafe-mode\u201d (only repay debt, only de-risk, no new exposure)<\/li>\n<\/ul>\n<p>Teams that treat monitoring as a product feature\u2014not a dashboard\u2014are the ones I take seriously.<\/li>\n<\/ul>\n<p>More builders are circling these exact layers than most people realize. I\u2019ve been tracking the narrative breadcrumbs here:<\/p>\n<p><a href=\"https:\/\/x.com\/Luckyman6886\/status\/2021217891140747512\" target=\"_blank\" rel=\"noopener\">Luckyman6886<\/a> \u2022 <a href=\"https:\/\/x.com\/WebThreeAI\/status\/2021062544488841676\" target=\"_blank\" rel=\"noopener\">WebThreeAI<\/a> \u2022 <a href=\"https:\/\/x.com\/Captain_1Kenobi\/status\/2021140855470555178\" target=\"_blank\" rel=\"noopener\">Captain_1Kenobi<\/a> \u2022 <a href=\"https:\/\/x.com\/ItsBitcoinWorld\/status\/2021141815865508074\" target=\"_blank\" rel=\"noopener\">ItsBitcoinWorld<\/a><\/p>\n<h3>DeFi use cases likely to grow first (next 6 months)<\/h3>\n<p>I\u2019m not interested in \u201cAI will trade better than everyone\u201d hype. The near-term winners are the boring-sounding automations that remove friction while keeping humans in control of goals and risk.<\/p>\n<ul>\n<li><strong>Auto-rebalancing vaults that react to market regimes<\/strong>Think rules like: reduce volatility exposure when funding flips; increase stables allocation when correlation spikes; rebalance when drift exceeds X%. The agent doesn\u2019t need to predict the future\u2014it needs to execute <strong>consistent policy<\/strong> with discipline.<\/li>\n<li><strong>Smarter liquidation protection and collateral management<\/strong>We\u2019ve watched liquidation cascades punish \u201cset-and-forget\u201d borrowers for years. An agent that monitors health factors and does small, frequent actions (repay a bit, swap collateral, hedge delta) can reduce forced liquidations\u2014if it\u2019s constrained and uses safe routing.<\/li>\n<li><strong>Intent-based swapping that reduces MEV pain<\/strong>Most users don\u2019t care about the trade path\u2014they care about the outcome. Intents + solvers can improve execution, but the design has to be resistant to the \u201cpredictable agent\u201d problem (because predictable flow is a magnet for extraction).<\/li>\n<li><strong>AI-assisted market making &amp; inventory management (with guardrails)<\/strong>Not \u201cblack-box bot prints money.\u201d More like: keep spreads within bounds, cap inventory, cut risk when volatility spikes, and stop when slippage or oracle deviation exceeds limits.<\/li>\n<li><strong>Credit\/underwriting primitives using onchain behavior signals<\/strong>This is powerful and dangerous. Onchain behavior can signal repayment patterns, risk tolerance, and trading style\u2014but it also creates privacy and bias issues fast. If teams can\u2019t explain the data, the incentives, and the failure modes, I treat it as a red flag.<\/li>\n<li><strong>Treasury automation for DAOs and onchain businesses<\/strong>Simple wins: automated streaming, yield routing with limits, rebalancing across stable pools, policy-driven diversification\u2014without a weekly governance vote for every minor move.<\/li>\n<\/ul>\n<h3>Where the new \u201cbillions in value\u201d could come from (without hand-waving)<\/h3>\n<p>\u201cBillions\u201d isn\u2019t a magic number. It has to show up in measurable flows. Here are the concrete drivers I\u2019m watching:<\/p>\n<ul>\n<li><strong>More users onboard because complexity drops<\/strong>If an agent can handle safe approvals, routing, and monitoring, the user experience starts to feel less like piloting a cockpit and more like setting rules.<\/li>\n<li><strong>Higher onchain volume because agents act more frequently than humans<\/strong>Humans rebalance monthly. Agents can rebalance when thresholds trigger. More small actions can mean more aggregate volume\u2014<em>if fees stay low enough on L2s<\/em> and execution is protected.<\/li>\n<li><strong>New fees: automation tooling, risk vaults, agent marketplaces<\/strong>Expect fee models around guardrails: simulation services, policy engines, solver networks, monitoring, insurance wrappers. The value isn\u2019t \u201cAI\u201d; it\u2019s <strong>reliable automation<\/strong>.<\/li>\n<li><strong>More capital efficiency<\/strong>Better routing and better hedging can reduce dead capital. Fewer bad liquidations can reduce system losses. This is where \u201cAI + onchain constraints\u201d can create real efficiency rather than just new speculation.<\/li>\n<\/ul>\n<p>The \u201cif\u201d conditions matter. If agents are unsafe, they won\u2019t scale. If intents are captured by extractive middlemen, users won\u2019t stick. If oracles remain fragile at the edges, agents become exploit accelerants.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-6351\" src=\"https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/Infra-plays-to-watch-what-needs-to-improve-for-the-thesis-to-hold.png\" alt=\"Infra plays to watch what needs to improve for the thesis to hold\" width=\"1536\" height=\"1024\" srcset=\"https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/Infra-plays-to-watch-what-needs-to-improve-for-the-thesis-to-hold.png 1536w, https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/Infra-plays-to-watch-what-needs-to-improve-for-the-thesis-to-hold-300x200.png 300w, https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/Infra-plays-to-watch-what-needs-to-improve-for-the-thesis-to-hold-1024x683.png 1024w, https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/Infra-plays-to-watch-what-needs-to-improve-for-the-thesis-to-hold-768x512.png 768w\" sizes=\"auto, (max-width: 1536px) 100vw, 1536px\" \/><\/p>\n<h3>Infra plays to watch: what needs to improve for the thesis to hold<\/h3>\n<p>If you want early signals (before the headlines), these are the things I track because they\u2019re the plumbing. When plumbing improves, products suddenly become possible.<\/p>\n<ul>\n<li><strong>Smart accounts \/ account abstraction UX<\/strong>: session keys, recovery, permission dashboards that normal people can understand.<\/li>\n<li><strong>Intents and solvers becoming mainstream<\/strong>: visible in major wallets\/DEX front ends, not just niche power-user tools.<\/li>\n<li><strong>Provable compute \/ verification tooling<\/strong>: even partial proofs and attestations help (the goal is reducing \u201ctrust me\u201d surfaces).<\/li>\n<li><strong>Onchain risk tooling<\/strong>: real-time exposures, stress tests, transparent strategy constraints, better health monitoring across venues.<\/li>\n<li><strong>Secure agent permissioning<\/strong>: revocation that works instantly, role-based controls, policy engines that can\u2019t be bypassed by prompt tricks.<\/li>\n<\/ul>\n<h3>The big risks (the part most threads skip)<\/h3>\n<p>If you automate finance, you don\u2019t just increase speed\u2014you increase the <strong>blast radius<\/strong>. Here are the failure modes I\u2019m treating as non-negotiable to address:<\/p>\n<ul>\n<li><strong>Prompt injection and tool misuse<\/strong>If an agent reads untrusted content (a webpage, a \u201chelpful\u201d message, even a token description) and that content influences tool calls, you can end up signing something you never intended. The fix is boring but effective: <strong>tool call allowlists<\/strong>, <strong>explicit transaction templates<\/strong>, and policy checks that operate outside the model.<\/li>\n<li><strong>Oracle manipulation + thin liquidity traps<\/strong>Attackers love predictable behaviors. If an agent buys whenever indicator X triggers, someone will try to manufacture X. Robust systems cross-check feeds, cap slippage, refuse to trade on suspicious data, and favor deeper liquidity.<\/li>\n<li><strong>MEV and sandwiching amplified by predictable agents<\/strong>MEV isn\u2019t new. The academic and builder communities have documented it for years. The twist here is that agents can become <em>more readable than humans<\/em>. If your agent always rebalances at the same thresholds, someone can camp those levels. Serious teams randomize timing, use protected routes, and avoid broadcasting fragile paths.<\/li>\n<li><strong>Model hallucinations turning into real losses<\/strong>A model confidently making up a detail is funny in a chatbot. It\u2019s expensive in a wallet. That\u2019s why I want agents that operate on <strong>verified state<\/strong> and <strong>restricted actions<\/strong>, not free-form \u201creasoning\u201d that can improvise transactions.<\/li>\n<li><strong>Regulatory headaches<\/strong>Autonomous financial agents can look like unlicensed portfolio management, especially if they custody funds or market \u201creturns.\u201d The safer direction is user-controlled permissioning, clear disclosures, and designs where the user sets policy and retains final authority (at least above thresholds).<\/li>\n<\/ul>\n<p>The teams I respect don\u2019t pretend these risks don\u2019t exist. They build like they expect failure: constraints first, simulation by default, transparency always, audits that cover the agent + contracts + offchain services\u2014not just one piece.<\/p>\n<h3>Investor angle: my checklist for separating real AI-crypto from buzzwords<\/h3>\n<p>I filter \u201cAI + DeFi\u201d projects with questions that sound simple but eliminate most of the fluff instantly:<\/p>\n<ul>\n<li><strong>Is it actually executing onchain?<\/strong> Or is it just a chatbot glued to a portfolio tracker?<\/li>\n<li><strong>Is there a clear permission model?<\/strong> Limits, roles, revocation, session duration, whitelists.<\/li>\n<li><strong>Can actions be reproduced and audited?<\/strong> Logs, signed intents, deterministic transaction building, transparent policies.<\/li>\n<li><strong>Where does the data come from?<\/strong> What happens if the oracle is wrong, delayed, or manipulated?<\/li>\n<li><strong>Are incentives aligned?<\/strong> Bonding, slashing, reputation, insurance, solver competition that benefits the user.<\/li>\n<li><strong>Does it have distribution?<\/strong> Wallet integrations, DEX integrations, real users\u2014or only timelines and token teasers?<\/li>\n<\/ul>\n<p>If a team can\u2019t answer these cleanly, I don\u2019t care how fancy the demo looks.<\/p>\n<h3>People also ask<\/h3>\n<ul>\n<li><strong>How can Ethereum be used for AI agents?<\/strong>As a settlement and coordination layer: payments, identity\/permissions via smart accounts, escrow and bonding, dispute rules, and audit trails. The AI thinks offchain, but Ethereum makes its actions accountable.<\/li>\n<li><strong>What is an onchain AI agent?<\/strong>Usually it\u2019s not a model running onchain. It\u2019s an offchain agent that uses an onchain wallet (often a smart account) plus a set of tools and constraints to execute transactions under strict policies.<\/li>\n<li><strong>Will AI replace DeFi traders?<\/strong>It\u2019ll automate a lot of strategies and execution. Humans will still set objectives, constraints, and risk budgets. The edge shifts from clicking fast to designing better policies and better protection.<\/li>\n<li><strong>Is AI DeFi safe?<\/strong>It can be <em>safer than manual<\/em> if it has guardrails (caps, simulations, whitelists, protected execution, monitoring). It can also be dramatically more dangerous if it\u2019s a black box with broad permissions.<\/li>\n<li><strong>What tokens benefit from AI + ETH?<\/strong>I don\u2019t treat this as a \u201cbuy list.\u201d I look at categories that capture value if adoption happens: smart account infrastructure, intent\/solver ecosystems, risk automation tooling, and protocols that become the default execution venues for automated flow. Then I ask: where are the fees, and who owns the user relationship?<\/li>\n<\/ul>\n<h3>Threads and sources I\u2019m using to track the narrative (quick references)<\/h3>\n<ul>\n<li><a href=\"https:\/\/x.com\/lucianlampdefi\/status\/2021202754329584078\" target=\"_blank\" rel=\"noopener\">https:\/\/x.com\/lucianlampdefi\/status\/2021202754329584078<\/a><\/li>\n<li><a href=\"https:\/\/x.com\/javliscom\/status\/2021412008256491944\" target=\"_blank\" rel=\"noopener\">https:\/\/x.com\/javliscom\/status\/2021412008256491944<\/a><\/li>\n<li><a href=\"https:\/\/x.com\/ThCryptoCook\/status\/2021017173137949138\" target=\"_blank\" rel=\"noopener\">https:\/\/x.com\/ThCryptoCook\/status\/2021017173137949138<\/a><\/li>\n<li><a href=\"https:\/\/x.com\/Tabl4me\/status\/2021230876563013795\" target=\"_blank\" rel=\"noopener\">https:\/\/x.com\/Tabl4me\/status\/2021230876563013795<\/a><\/li>\n<li><a href=\"https:\/\/x.com\/houseofai_swan\/status\/2021276364062757069\" target=\"_blank\" rel=\"noopener\">https:\/\/x.com\/houseofai_swan\/status\/2021276364062757069<\/a><\/li>\n<li><a href=\"https:\/\/x.com\/BSCNews\/status\/2021073501130719374\" target=\"_blank\" rel=\"noopener\">https:\/\/x.com\/BSCNews\/status\/2021073501130719374<\/a><\/li>\n<li><a href=\"https:\/\/x.com\/Luckyman6886\/status\/2021217891140747512\" target=\"_blank\" rel=\"noopener\">https:\/\/x.com\/Luckyman6886\/status\/2021217891140747512<\/a><\/li>\n<li><a href=\"https:\/\/x.com\/WebThreeAI\/status\/2021062544488841676\" target=\"_blank\" rel=\"noopener\">https:\/\/x.com\/WebThreeAI\/status\/2021062544488841676<\/a><\/li>\n<li><a href=\"https:\/\/x.com\/Captain_1Kenobi\/status\/2021140855470555178\" target=\"_blank\" rel=\"noopener\">https:\/\/x.com\/Captain_1Kenobi\/status\/2021140855470555178<\/a><\/li>\n<li><a href=\"https:\/\/x.com\/ItsBitcoinWorld\/status\/2021141815865508074\" target=\"_blank\" rel=\"noopener\">https:\/\/x.com\/ItsBitcoinWorld\/status\/2021141815865508074<\/a><\/li>\n<li><a href=\"https:\/\/x.com\/0xshai\/status\/2021147938236531180\" target=\"_blank\" rel=\"noopener\">https:\/\/x.com\/0xshai\/status\/2021147938236531180<\/a><\/li>\n<li><a href=\"https:\/\/x.com\/VitalikButerin\/status\/2020963864175657102\" target=\"_blank\" rel=\"noopener\">https:\/\/x.com\/VitalikButerin\/status\/2020963864175657102<\/a><\/li>\n<\/ul>\n<p><strong>Now the uncomfortable question:<\/strong> if this is truly going to move from \u201cthreads\u201d to \u201creal usage,\u201d what would I expect to see <em>change week by week<\/em> across wallets, DEXs, and L2s\u2014before the market fully prices it in?<\/p>\n<p>That\u2019s what I\u2019m laying out next.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-6350\" src=\"https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/What-I-expect-over-the-next-6-months-and-what-Im-watching-weekly.png\" alt=\"What I expect over the next 6 months (and what I\u2019m watching weekly)\" width=\"1536\" height=\"1024\" srcset=\"https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/What-I-expect-over-the-next-6-months-and-what-Im-watching-weekly.png 1536w, https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/What-I-expect-over-the-next-6-months-and-what-Im-watching-weekly-300x200.png 300w, https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/What-I-expect-over-the-next-6-months-and-what-Im-watching-weekly-1024x683.png 1024w, https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/What-I-expect-over-the-next-6-months-and-what-Im-watching-weekly-768x512.png 768w\" sizes=\"auto, (max-width: 1536px) 100vw, 1536px\" \/><\/p>\n<h2>What I expect over the next 6 months (and what I\u2019m watching weekly)<\/h2>\n<p>If this AI + Ethereum thing is going to matter, it won\u2019t show up first as \u201cAI tokens pumping.\u201d It\u2019ll show up as boring product wins: safer permissions, more intent-based trades, and DeFi protocols quietly shipping agent-friendly rails that normal people never have to think about.<\/p>\n<p>So here\u2019s my grounded framework for the next six months. I\u2019m watching four lanes every week:<\/p>\n<ul>\n<li><strong>Wallets:<\/strong> smart accounts, session keys, spending policies, recovery<\/li>\n<li><strong>DeFi protocols:<\/strong> \u201cagent-ready\u201d flows, guardrails, better APIs<\/li>\n<li><strong>Agent frameworks:<\/strong> tooling that turns strategies into accountable execution<\/li>\n<li><strong>Verification\/security:<\/strong> audits, simulations, attestations, monitoring becoming default<\/li>\n<\/ul>\n<p>My base case: we get a wave of <em>small but meaningful<\/em> UX shifts that make automation feel less scary. The bull case: one or two mainstream wallets make \u201csafe automation\u201d a one-click feature and volumes follow. The bear case: a few high-profile agent blowups (or \u201cAI-managed vault\u201d incidents) scare users back into manual mode.<\/p>\n<blockquote><p><strong>The tell:<\/strong> the winning products won\u2019t market \u201cAI.\u201d They\u2019ll market <em>guarantees<\/em>\u2014limits, reversibility, transparency, and proof you can audit.<\/p><\/blockquote>\n<hr \/>\n<h3>The \u201cif this is real\u201d scoreboard: metrics that should move<\/h3>\n<p>I don\u2019t need a thousand new narratives. I need a handful of metrics to trend in the right direction. Here\u2019s the scoreboard I\u2019m tracking weekly (and yes, I\u2019ll be linking dashboards when I spot good ones):<\/p>\n<ul>\n<li><strong>Smart-account adoption (account abstraction):<\/strong> a steady rise in smart accounts used by real humans\u2014not just airdrop farmers. The easiest proxy is growth in ERC-4337-style user operations and the number of unique accounts interacting with them.<em>What would convince me:<\/em> major consumer wallets pushing smart accounts by default (with clear recovery), and more apps treating smart accounts as the \u201cnormal\u201d wallet type.<\/li>\n<li><strong>Session keys + policy permissions becoming normal:<\/strong> more apps offering \u201capprove a policy\u201d instead of \u201capprove unlimited token spend.\u201d<em>Real-world example of what I want to see:<\/em> \u201cAllow this bot to rebalance up to $50\/day, only between USDC\/ETH, only on these DEXs, and revoke anytime.\u201d If I can\u2019t set those rules, I\u2019m not letting an agent touch my funds.<\/li>\n<li><strong>Intent-based trade share rising:<\/strong> more swaps routed through intent systems (where you specify the outcome, solvers compete to fill it, and you get MEV protection-like benefits).<em>What would convince me:<\/em> intent flow becoming the default in popular aggregators and wallets for regular swaps, not just power-user features. (Think of systems like CoW Swap-style batch auctions, or UniswapX-style routing\u2014different implementations, same direction.)<\/li>\n<li><strong>Automated vault volume with transparent rules:<\/strong> assets in vaults\/strategies that publish their logic and constraints clearly (rebalance rules, risk limits, slippage caps, allowed venues).<em>What would convince me:<\/em> vaults that show \u201cwhy\u201d a trade happened and provide a readable action log, not just \u201ctrust the black box.\u201d<\/li>\n<li><strong>Protocols shipping agent-friendly surfaces:<\/strong> lending\/perps\/DEX protocols adding official automation hooks: safe callbacks, intent endpoints, simulation tooling, and guardrail modules (rate limits, circuit breakers, configurable constraints).<em>What would convince me:<\/em> release notes that mention agent safety explicitly\u2014because that\u2019s where real product maturity shows up.<\/li>\n<li><strong>Onchain attestations and proofs showing up in normal UX:<\/strong> not full sci-fi \u201cproof of everything,\u201d but practical attestations like \u201cthis action was simulated,\u201d \u201cthis execution met constraints,\u201d \u201cthis agent posted a bond,\u201d or \u201cthis vault has continuous monitoring.\u201d<em>What would convince me:<\/em> standardized receipts\/log formats for agent actions, so monitoring tools can plug in easily.<\/li>\n<\/ul>\n<p>One thing I\u2019m <strong>not<\/strong> counting as traction: \u201cWe added an AI chat tab to the app.\u201d Cool, but that\u2019s not the hard part. The hard part is safe execution under constraints.<\/p>\n<p>Also, a reminder from the last couple years: losses don\u2019t come from a lack of intelligence, they come from a lack of controls. Reports from groups like <a href=\"https:\/\/www.immunefi.com\/research\/\" target=\"_blank\" rel=\"noopener\">Immunefi\u2019s research<\/a> and <a href=\"https:\/\/www.chainalysis.com\/reports\/\" target=\"_blank\" rel=\"noopener\">Chainalysis reports<\/a> have consistently shown that exploits, bad permissions, and operational mistakes are where users get hit. If AI increases the speed of decision-making, controls have to improve even faster.<\/p>\n<hr \/>\n<h3><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-6349\" src=\"https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/How-Id-approach-it-as-a-reader-participate-without-getting-wrecked.png\" alt=\"How I\u2019d approach it as a reader participate without getting wrecked\" width=\"1536\" height=\"1024\" srcset=\"https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/How-Id-approach-it-as-a-reader-participate-without-getting-wrecked.png 1536w, https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/How-Id-approach-it-as-a-reader-participate-without-getting-wrecked-300x200.png 300w, https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/How-Id-approach-it-as-a-reader-participate-without-getting-wrecked-1024x683.png 1024w, https:\/\/cryptolinks.com\/news\/wp-content\/uploads\/2026\/02\/How-Id-approach-it-as-a-reader-participate-without-getting-wrecked-768x512.png 768w\" sizes=\"auto, (max-width: 1536px) 100vw, 1536px\" \/><\/h3>\n<h3>How I\u2019d approach it as a reader: participate without getting wrecked<\/h3>\n<p>If you want exposure to this trend without becoming a test dummy, here\u2019s how I\u2019d do it (this is basically my personal rulebook):<\/p>\n<ul>\n<li><strong>Start tiny and earn the right to scale.<\/strong> I treat the first 2\u20134 weeks as paid learning. If a strategy can\u2019t explain itself clearly at $50, it doesn\u2019t deserve $5,000.<\/li>\n<li><strong>Prefer non-custodial automation.<\/strong> If the \u201cagent\u201d needs you to deposit into a wallet you don\u2019t control, you\u2019re not testing automation\u2014you\u2019re taking counterparty risk.<\/li>\n<li><strong>Use strict permissions like it\u2019s a religion.<\/strong> If you can\u2019t set:\n<ul>\n<li>daily\/weekly spend caps<\/li>\n<li>allowlisted contracts\/protocols<\/li>\n<li>token allowlists<\/li>\n<li>slippage ceilings<\/li>\n<li>time-boxed session keys (auto-expire)<\/li>\n<li>one-click revoke<\/li>\n<\/ul>\n<p>\u2026then you\u2019re basically giving an unknown system a blank check.<\/li>\n<li><strong>Pick strategies you can describe in one sentence.<\/strong> Example: \u201cRebalance 80\/20 ETH\/USDC weekly if volatility spikes, otherwise monthly.\u201d If the pitch sounds like \u201cour AI finds alpha across chains,\u201d I assume it\u2019s either overfit, opaque, or both.<\/li>\n<li><strong>Demand simulations and readable logs.<\/strong> The best teams will show:\n<ul>\n<li>pre-trade simulation (expected outcome range)<\/li>\n<li>post-trade receipt (what happened vs expected)<\/li>\n<li>reason codes (why the agent acted)<\/li>\n<li>alerting (Telegram\/email\/onchain watcher) when constraints are close to breaking<\/li>\n<\/ul>\n<\/li>\n<li><strong>Watch for \u201cMEV-shaped\u201d pain.<\/strong> If an agent trades predictably (same time every day, same venue, same sizing), it can become an easy target. I prefer systems that randomize timing, use intent-based execution, and avoid shouting their next move to the mempool.<\/li>\n<li><strong>Assume models fail.<\/strong> Not \u201cmight.\u201d They will. The right question is: <em>what happens when it fails?<\/em> I look for circuit breakers (pause conditions), safe-mode defaults, and the ability to degrade gracefully to \u201cdo nothing.\u201d<\/li>\n<li><strong>Check audits, but also check what audits don\u2019t cover.<\/strong> Audits matter. But the biggest AI-specific risk is often the tool layer: bad transaction building, unsafe allowances, prompt injection into web-connected agents, or a malicious \u201cdata source\u201d telling the agent a lie. I want to see threat models and post-mortems, not just badges.<\/li>\n<\/ul>\n<p>My personal red flag: any project that hides behind \u201cproprietary AI\u201d while also asking for broad permissions. In DeFi, opacity plus authority is how people get wrecked.<\/p>\n<hr \/>\n<h3>A quick wrap: what matters from here<\/h3>\n<p>Right now, this space is shifting from \u201ccool demo\u201d to something that actually has teeth: coordination + execution + accountability. That\u2019s the real meaning of what Vitalik\u2019s framing kicked up\u2014Ethereum isn\u2019t where the AI \u201cthinks,\u201d it\u2019s where the AI <em>commits<\/em> actions in a way other systems can verify and build on.<\/p>\n<p>The opportunity is real, but I\u2019m betting the winners look a little boring on the surface:<\/p>\n<ul>\n<li><strong>Security as a feature<\/strong> (not a disclaimer)<\/li>\n<li><strong>Permissions as a product<\/strong> (not a developer setting)<\/li>\n<li><strong>Verification and logs<\/strong> that make automation auditable<\/li>\n<\/ul>\n<p>I\u2019ll keep posting updates and dashboards as the metrics start moving on <a href=\"https:\/\/cryptolinks.com\/news\/\" target=\"_blank\" rel=\"noopener\">https:\/\/cryptolinks.com\/news\/<\/a>. Next up, I\u2019m putting together a short, practical shortlist: the categories I think are most likely to break out, and the red flags that usually show up right before users become exit liquidity.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>DeFi is pro mode\u2014one wrong click can wreck you. I show how Vitalik\u2019s ETH plan lets AI agents on Ethereum + L2s act with limits, intent flows, and onchain logs.<\/p>\n","protected":false},"author":1,"featured_media":6354,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[],"class_list":["post-6343","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-uncategorized"],"_links":{"self":[{"href":"https:\/\/cryptolinks.com\/news\/wp-json\/wp\/v2\/posts\/6343","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/cryptolinks.com\/news\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/cryptolinks.com\/news\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/cryptolinks.com\/news\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/cryptolinks.com\/news\/wp-json\/wp\/v2\/comments?post=6343"}],"version-history":[{"count":5,"href":"https:\/\/cryptolinks.com\/news\/wp-json\/wp\/v2\/posts\/6343\/revisions"}],"predecessor-version":[{"id":6355,"href":"https:\/\/cryptolinks.com\/news\/wp-json\/wp\/v2\/posts\/6343\/revisions\/6355"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/cryptolinks.com\/news\/wp-json\/wp\/v2\/media\/6354"}],"wp:attachment":[{"href":"https:\/\/cryptolinks.com\/news\/wp-json\/wp\/v2\/media?parent=6343"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/cryptolinks.com\/news\/wp-json\/wp\/v2\/categories?post=6343"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/cryptolinks.com\/news\/wp-json\/wp\/v2\/tags?post=6343"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}