Will AI Replace Crypto Traders by 2028? Nansen CEO Makes a Bold Prediction
2026/07/31 11:19:00
Crypto markets could contain more AI trading agents than human traders by 2028, according to Nansen CEO and co-founder Alex Svanevik. The prediction reflects the rapid development of agentic trading systems that can analyze market information, interpret broader instructions and perform financial actions with varying levels of independence. However, having more agents than humans would not prove that artificial intelligence consistently generates better returns or that human participation is disappearing.The more important question is which parts of AI in crypto trading can be performed reliably. Research, portfolio monitoring and routine execution may become increasingly automated, while strategy selection, accountability and responses to unusual market events remain more difficult to delegate. The transition could improve access to sophisticated trading tools, but it also introduces concerns involving profitability, manipulated data, wallet permissions, correlated trading and unclear responsibility when an autonomous system makes a costly mistake.
Nansen CEO Predicts AI Trading Agents Will Outnumber Human Traders by 2028
What Alex Svanevik Actually Predicted About AI Trading Agents
Nansen CEO and co-founder Alex Svanevik predicts that AI trading agents could outnumber human traders by 2028, as investors increasingly delegate market research, portfolio monitoring and trade execution to autonomous software. In a July 2026 interview, Svanevik said he would be surprised if the number of trading agents did not exceed the number of human traders within approximately two years. His forecast builds on Nansen’s earlier agentic trading forecast, which suggested that agents could become the default way people interact with investment markets by 2028. Instead of manually examining charts, tracking hundreds of wallets and approving every individual transaction, users could define an investment objective, risk tolerance and permitted assets while an AI agent manages more of the underlying workflow.
However, Svanevik’s statement should not be interpreted as a confirmed prediction that AI will eliminate every crypto trader or automatically outperform human investors. He specifically referred to the number of AI trading agents, not their profitability, trading volume or share of total market activity. One investor, institution, protocol or fund could deploy multiple specialized agents for research, execution, portfolio rebalancing and risk monitoring, allowing the agent population to expand much faster than the human user base. The number of active agents could therefore exceed the number of people using them without proving that machines have achieved superior investment judgment.
How Nansen Is Building Toward Agentic Crypto Trading
Nansen’s product strategy helps explain why Svanevik expects agentic trading to grow quickly. The company is evolving from a platform focused primarily on onchain analytics into an integrated research-and-execution environment where users can identify market signals and prepare trades without moving between several applications. Nansen says its infrastructure includes more than 500 million labeled blockchain addresses, providing additional context about exchanges, funds, whales, profitable traders and other onchain entities. Its AI tools can use this information to examine wallet behavior, token-holder distribution, decentralized exchange flows and historical performance before proposing an action.
Nansen has also integrated spot trading and Hyperliquid perpetual trading, expanding its strategy beyond conventional cryptocurrency analysis. Svanevik reported that the broader trading feature had processed more than $500 million in cumulative volume, although this company-reported figure represents platform activity rather than independently verified autonomous-agent volume or profit. Higher-autonomy agents are still being evaluated through backtesting and paper trading, and the economics require improvement. Svanevik disclosed that one experimental agent produced only $23 in profit while consuming approximately $700 in AI inference costs, illustrating the gap between technical capability and commercially sustainable performance.
Why AI Trading Agents Could Transform Crypto and Onchain Trading
Crypto markets are particularly suited to crypto AI agents because they operate continuously and produce large quantities of public, machine-readable data. Blockchain transactions are timestamped and generally observable in real time, allowing an agent to analyze token transfers, decentralized exchange liquidity, lending positions, bridge activity, governance events and changes in collateral across multiple protocols. By connecting this information with price feeds, market sentiment and risk parameters, an agent could complete a multistage workflow that would be difficult for one person to manage manually. It could detect an emerging market signal, evaluate liquidity and counterparty exposure, compare execution routes, calculate an appropriate position size and place or adjust a trade through smart contracts. It could then monitor slippage, funding rates, collateral requirements and liquidation thresholds continuously, reducing the delay between discovering information and acting on it.
Wider use of agentic crypto trading could also change how liquidity, competition and portfolio management function across onchain markets. Retail users may gain access to automated research and risk-monitoring capabilities that previously required analysts, quantitative tools and dedicated execution systems, while professional firms could coordinate specialized agents for market making, arbitrage, hedging and cross-chain treasury management. Agents could divide large orders across several venues, rebalance portfolios when exposure exceeds predefined limits and move capital between spot, derivatives, lending and liquidity pools as conditions change. This would make trading more responsive and personalized, but it would also shift the source of competitive advantage. Instead of relying mainly on faster manual decisions, market participants would compete through data quality, execution infrastructure, model design, secure wallet permissions and disciplined risk controls.
Will AI Replace Crypto Traders? Benefits, Limitations and Risks by 2028
Whether AI will replace crypto traders by 2028 is not a simple choice between complete automation and manual trading. AI systems may take responsibility for a growing share of market analysis and decision-making, but widespread adoption would not establish that they can consistently outperform experienced traders. Their long-term role will depend on measurable performance, security, accountability and their ability to respond safely when markets behave differently from the historical conditions used to develop them.
Benefits of AI Trading Agents for Crypto Traders
The main benefit of AI trading agents is their ability to apply a structured process consistently across numerous assets and strategies. Human traders can become distracted, abandon risk rules after losses or make impulsive decisions during periods of extreme volatility, while a properly configured agent can evaluate each opportunity using the same position-sizing criteria, exposure limits and exit conditions. Unlike conventional crypto trading bots, agentic systems may interpret broader objectives and adjust their analysis as conditions change instead of relying only on fixed instructions. Natural-language tools may also make these capabilities more accessible to users without advanced programming knowledge by translating objectives such as reducing portfolio concentration, controlling drawdowns or maintaining a specific asset allocation into measurable conditions. These systems cannot guarantee higher returns, but they could improve trading discipline, document decisions more clearly and reduce errors caused by fatigue, inconsistent analysis or emotional reactions.
Why AI Trading Agents Cannot Guarantee Profits
Strong analytical capabilities do not automatically produce profitable crypto trades because financial markets continually change. A strategy that performs well during a sustained uptrend may fail when prices become range-bound, liquidity weakens or correlations between assets suddenly shift. AI models can mistake correlation for causation, rely too heavily on recent patterns or provide convincing explanations for signals with limited predictive value. Backtests can also produce misleading results when a strategy is repeatedly adjusted to fit past prices or when training data indirectly contains information from the evaluation period. The official CFTC AI warning emphasizes that artificial intelligence cannot predict the future or unexpected market changes and should not be associated with guaranteed-return claims. Reliable evaluation therefore requires out-of-sample testing across bullish, bearish and sideways markets, with performance measured through risk-adjusted returns, maximum drawdown and consistency rather than headline profits alone.
Security and Operational Risks of Autonomous Crypto Trading
Autonomous crypto trading introduces serious security and operational risks when an AI agent receives permission to interact with wallets, exchanges or smart contracts. A malicious instruction hidden in a website, social-media post or token description could influence an agent that processes untrusted information, while manipulated price feeds, compromised APIs and inaccurate market data could trigger trades under false conditions. An agent might also misunderstand an open-ended objective, interact with an unsafe contract or continue executing a strategy after market conditions or the user’s intentions have changed. The potential damage depends on its permissions, making dedicated accounts, asset allowlists, withdrawal restrictions, position limits, manual approval thresholds and immediate shutdown controls essential. Detailed activity logs should show what information the agent used, which actions it performed and whether it remained within its authorized scope. Without these protections, the speed of autonomous execution could transform a small analytical mistake into multiple irreversible transactions before the user can intervene.
Market, Regulatory and Human-Trader Outlook for 2028
The large-scale adoption of AI trading agents could affect liquidity, volatility and regulation across crypto markets. If many agents rely on similar models, datasets and strategies, they may respond to the same event by buying or selling simultaneously, producing crowded positions, liquidation cascades or deeper liquidity shortages. Automated agents could also compete for identical arbitrage opportunities, increase network congestion and expose transactions to front-running or other forms of manipulation. Regulators may consequently require stronger disclosures, system testing, record-keeping, suitability controls and clear responsibility for losses caused by autonomous decisions. Questions will remain about whether accountability belongs to the user, model developer, data provider, wallet service or execution platform when an agent acts unexpectedly. By 2028, AI may automate substantial parts of crypto trading, but traders who can evaluate model outputs, recognize changing market conditions and impose effective risk controls are likely to remain important.
Conclusion: Will AI Replace Human Crypto Traders by 2028?
AI trading agents are likely to become more visible across crypto markets by 2028, but their growth should not be confused with proven superiority over human traders. The technology is well suited to automating repetitive research, applying predefined rules and coordinating complex digital-asset workflows, yet reliable profitability remains difficult to demonstrate after model costs, execution assumptions and changing market conditions are considered. Security failures, manipulated inputs, correlated strategies and unresolved accountability could also limit how much capital users are willing to delegate. The most plausible outcome is therefore a hybrid market in which AI performs more analytical and operational tasks while people remain responsible for selecting objectives, validating unusual decisions and accepting legal and financial responsibility. Rather than simply replacing traders, AI may redefine what successful crypto trading requires by making model supervision, data quality and automated-risk governance increasingly important sources of competitive advantage.
|
KuCoin is celebrating its 9th anniversary with a special platform campaign filled with exclusive rewards, trading activities, and limited-time offers. Don’t miss the chance to participate and enjoy the benefits as the exchange marks nine years of growth and innovation. Visit the official campaign page now:
|
Frequently Asked Questions
How is an AI trading agent different from an AI analysis tool?
An AI analysis tool summarizes market information and suggests possible actions, but the user still makes and executes the final decision. An AI trading agent can move beyond recommendations by placing orders, adjusting positions or rebalancing a portfolio within its assigned permissions. Some platforms use the term “agent” for tools that cannot act independently, so users should verify exactly what the system can access and execute.
How should an AI trading agent’s performance be evaluated?
Performance should be measured using net returns after transaction costs, funding payments, slippage and computing expenses. Users should also examine maximum drawdown, risk-adjusted returns, consistency across different market conditions and performance against simple benchmarks such as buy-and-hold. Results from live or out-of-sample testing are more informative than backtests built around historical data the model may have already encountered.
Does an AI trading agent need access to private keys?
A responsibly designed agent should not require users to disclose a wallet seed phrase or unrestricted private key. Safer systems use limited API permissions, smart accounts or separately funded wallets that restrict withdrawals, approved assets and maximum transaction values. Users should never enter recovery phrases into an AI chatbot, browser extension or unverified trading application, regardless of its claimed performance.
Can AI trading agents operate across multiple blockchains?
An agent can work across multiple networks if it is connected to compatible wallets, bridges, decentralized exchanges and data providers. Cross-chain activity introduces additional risks involving bridge security, transaction finality, network congestion, liquidity fragmentation and changing gas costs. An agent should verify the destination network, token contract and available liquidity before moving funds or executing a cross-chain strategy.
Are trades made by AI agents subject to tax?
Using an AI agent generally does not remove the account owner’s tax obligations. Sales, swaps, rewards, derivatives settlements and transfers between different assets may create reportable or taxable events depending on the user’s jurisdiction. Because an agent may execute many transactions automatically, users should retain complete records showing timestamps, asset quantities, market values, fees and transaction hashes and consult a qualified tax professional where necessary.
Can an AI trading agent be proven completely safe?
No autonomous financial system can be guaranteed completely safe. Independent code audits, controlled permissions, adversarial testing, transaction simulations and emergency shutdown controls can reduce risk, but they cannot eliminate unexpected model behavior, software vulnerabilities or market losses. Users should also confirm whether the platform discloses its data sources, keeps detailed action logs and has a clear process for reporting unauthorized or incorrectly executed transactions.
Can traders manipulate an AI agent’s decisions?
AI agents may be influenced by fake news, coordinated social-media activity, misleading token information, manipulated price feeds or abnormal transactions in illiquid markets. Attackers could attempt to create signals that appear legitimate or insert malicious instructions into information processed by the model. Using several independent data sources, excluding low-liquidity assets and requiring confirmation for unusual transactions can reduce exposure to these tactics.
What is the safest way to test an AI crypto trading agent?
Beginners should start with paper trading or a sandbox environment before allowing an agent to use real funds. The next stage should involve a small, separately funded account with withdrawals disabled, low position limits, approved-asset restrictions and no or minimal leverage. Every action should be reviewed during the testing period, and access should be expanded only after the system demonstrates consistent compliance with the user’s instructions and risk limits.
How Much Does It Cost to Run an AI Crypto Trading Agent?
The cost of operating an AI crypto trading agent can include platform subscriptions, model-inference charges, premium market data, blockchain gas fees, trading commissions, slippage and funding payments on leveraged positions. More frequent analysis and execution may increase these expenses significantly, even when a strategy produces a positive gross return. Traders should therefore compare net performance after every operating and transaction cost rather than evaluating an agent solely by its reported profit or win rate.
What Should Traders Look for When Choosing an AI Trading Agent?
Traders should evaluate an AI trading agent based on transparent performance reporting, supported exchanges and blockchains, configurable risk limits, permission controls, independent security reviews and detailed transaction logs. A credible platform should clearly explain how its agent uses market data, protects user assets and responds when an API, model or data provider fails. Users should be cautious of services promising guaranteed returns, hiding fees or requesting seed phrases and unrestricted access to funds.
Disclaimer: This article is for informational purposes only and does not constitute financial advice. Crypto and AI-powered trading involve substantial risk, so always conduct independent research before investing.