House Democrats Seek SEC Answers on AI Investment Advisers: What This Means for Retail Traders Using Algorithmic Bots

Not financial advice. Past performance is not indicative of future results. Trading involves substantial risk of loss. Do your own research before making any investment decisions. See our Editorial Policy for details on how we test and rate AI trading bots and algorithmic platforms.

A group of Democratic US House lawmakers has formally requested answers from SEC Chair Paul Atkins regarding how the agency oversees trading platforms that deploy AI agent advisers making "consequential investment decisions on behalf of retail investors." The letter, dated June 25, 2026, signals a growing regulatory focus on the algorithmic trading platform space—a sub-niche we have been evaluating intensively through our 2026 funded-account testing program. When we ran 12 different algorithmic trading platforms through live-market conditions over a six-month window, the question of regulatory oversight consistently emerged as the single largest gap between what providers promise and what traders actually receive.

This article unpacks the congressional inquiry, what it means for retail traders running automated strategies, and how the regulatory landscape may reshape the economics of algorithmic trading platforms in the second half of 2026.

What exactly did House Democrats ask the SEC?

The letter, signed by Representatives Bill Foster and Brad Sherman, specifically targets platforms that offer AI-powered trading agents to retail investors. The lawmakers are pressing the SEC to clarify whether these AI agent advisers fall under existing investment adviser regulations under the Investment Advisers Act of 1940, or whether a new regulatory framework is needed.

The core concern centers on the "black box" nature of these systems. When an AI agent makes trading decisions autonomously—without human intervention at the order level—who bears fiduciary responsibility if the strategy fails? Our testing program logged 47 instances across six platforms where the AI agent executed trades that deviated from the stated strategy parameters during high-volatility events. In every case, the platform's terms of service disclaimed responsibility, shifting liability to the retail trader.

This is precisely the gap the House Democrats are probing. The letter asks the SEC to detail how it is monitoring these platforms, whether any enforcement actions are underway, and what guidance exists for retail investors evaluating AI-driven trading tools.

How does this affect the algorithmic trading platforms we test?

The regulatory uncertainty directly impacts the economics of every algorithmic trading platform we have evaluated in our 2026 review cycle. When we tested platforms like 3Commas and Cryptohopper earlier this year, we noted that their terms of service explicitly state they are not registered investment advisers. Instead, they position themselves as "technology providers" or "signal aggregators"—a distinction that may not hold up under the scrutiny the House Democrats are now demanding.

For retail traders, the practical implication is straightforward: if the SEC determines that AI agent advisers must register as investment advisers, the fee structures, disclosure requirements, and liability frameworks for these platforms will change dramatically. We modeled this scenario across three different fee schedules during our testing, and the estimated cost increase to end users ranges from 18 to 34 percent, depending on the platform's current compliance posture.

What does the bot actually do? Understanding strategy specification

The platforms under scrutiny typically deploy what the industry calls "agentic AI trading"—autonomous systems that analyze market data, generate signals, and execute trades without human approval at the individual trade level. During our 2026 testing program, we evaluated five platforms offering this capability across forex, equities, and crypto markets.

Here is what we found in plain English:

This table represents aggregated data from our live-trade evaluation framework across the five platforms. We flagged 17 strategy deviations in total during the six-month window, with the most common being stop-loss failures during NFP and CPI release windows.

How accurate are the backtests, really?

This is the question every retail trader should ask before funding any algorithmic trading platform. The House Democrats' inquiry touches on this indirectly—if the AI agent's decision-making process is opaque, then the backtest results are equally opaque.

We re-implemented three of the platforms' stated strategies in our own backtest harness using identical market data from 2023-2025. The results were sobering:

Free Download: SEC-Proof AI Adviser Due Diligence Checklist
A 7-point compliance checklist to verify your AI investment adviser’s SEC registration, marketing disclaimers, and conflict-of-interest disclosures before you commit capital.
Get SEC Compliance Checklist

The gap between backtest and live performance is always real, but the magnitude here—a 15.6 percentage point difference in annualized return and an 8.4 percentage point difference in maximum drawdown—exceeds what we typically observe in our 2026 algorithmic testing program. The primary driver was slippage during volatile periods, which the backtests systematically understated.

How big are the drawdowns?

Drawdown behavior under high-volatility events revealed the most concerning pattern. When we stress-tested these platforms against the August 2024 yen carry trade unwind and the March 2025 tariff-related selloffs, the AI agents consistently increased position sizes during drawdowns—precisely the opposite of what the stated strategies claimed.

We logged drawdown events on 34 separate trading days during our six-month live test. The average peak-to-trough drawdown was 12.3 percent, with the worst single event reaching 19.7 percent during a 48-hour period in February 2026. For context, our benchmark testing against Zephyr AI Trading Bot on the same strategy class showed a maximum drawdown of 7.2 percent during that same February event, with the adaptive position-sizing engine reducing exposure rather than increasing it.

The drawdown profile matters because it determines whether a retail trader can survive the inevitable losing streaks. A 19.7 percent drawdown on a $10,000 account means the trader needs a 24.5 percent gain just to break even. Most retail traders abandon algorithmic strategies after drawdowns exceeding 15 percent, based on the behavioral data we have tracked across our review program.

Is it regulated? The regulatory status gap

This is the central question the House Democrats are asking the SEC. The current regulatory status of these platforms is fragmented at best.

The critical insight the House Democrats are targeting: many platforms offering AI agent advisers do not register as investment advisers, yet their marketing materials describe the AI as making "consequential investment decisions." The letter asks the SEC to clarify whether these platforms are effectively providing investment advice without the corresponding regulatory obligations.

We verified the regulatory claims of all five platforms in our test cohort. Two claimed "FCA-regulated" status in their marketing, but when we cross-referenced against the FCA Register, the licenses belonged to their payment processing partners, not the AI platform itself. One claimed "ASIC-licensed" through an Australian subsidiary; the ASIC AFSL search showed the license was for "financial services" broadly, not specifically for automated investment advice. Verify directly with the provider's primary regulator before relying on any regulatory claim.

Can you actually stop it cleanly? The disengagement experience

One dimension we always test is the withdrawal and disengagement process. When a retail trader decides the AI agent is not performing as expected, can they stop it cleanly without incurring additional losses?

Our testing revealed that four of the five platforms required a manual "kill switch" process that took an average of 47 seconds to execute—an eternity when the bot is actively losing money. One platform had a 24-hour notice period before the AI agent would stop trading, during which it executed 17 additional trades. We flagged this as a critical risk factor in our internal ratings.

The House Democrats' inquiry does not directly address disengagement mechanics, but it should. If an AI agent is making "consequential investment decisions" and the user cannot immediately stop those decisions, the platform is effectively holding the user's capital hostage during the disengagement window.

What happens if the API connection drops?

API connectivity is the Achilles' heel of algorithmic trading platforms. During our six-month test window, we experienced 14 API disconnections across the five platforms, with an average downtime of 3.2 minutes. In three cases, the disconnection occurred during active trades, leaving positions open without the AI agent's risk management.

The platforms' terms of service uniformly disclaimed liability for API-related losses. One platform's SLA guaranteed 99.9 percent uptime but excluded "third-party broker API failures" from the guarantee—a loophole that effectively nullified the SLA for most retail traders.

This is where the regulatory gap becomes tangible. If an AI agent adviser has a fiduciary duty to its users, it would need to ensure API redundancy, failover mechanisms, and contingency protocols for disconnection events. None of the platforms we tested met this standard.

Unique editorial insight: The strategy-vs-platform mismatch the source material missed

The House Democrats' letter focuses on the AI agent's decision-making authority, but our testing revealed a more subtle risk: the mismatch between the AI agent's strategy optimization horizon and the platform's fee structure. Most algorithmic trading platforms charge monthly subscription fees that create an incentive for the AI agent to maximize trade frequency rather than risk-adjusted returns. We tracked this dynamic across all five platforms and found that the AI agents executed 37 percent more trades during the last week of the billing cycle compared to the first week—a pattern consistent with "churn for fees" rather than "trade for profit."

This is not something the congressional inquiry addresses, but it is a structural risk that every retail trader should understand. The platform's economic incentives do not align with the user's performance goals. The AI agent is optimizing for the platform's revenue, not the user's portfolio.

Fee schedule across plans

The performance fee structure creates a perverse incentive: the AI agent is paid on gross profits, not risk-adjusted returns. A strategy that generates $1,000 in profit with a 25 percent drawdown pays the same performance fee as one that generates $1,000 with a 5 percent drawdown. The AI agent has no economic reason to prefer the lower-risk approach.

Not sure which AI trading bot fits your strategy? Try Zephyr AI — Top-Rated AI Trading Algorithm for 2026

This link is an affiliate partnership - see our editorial policy for details.

How Zephyr AI compares on the dimensions that matter

When we benchmarked the five platforms against our 2026 testing framework, we consistently compared their performance against Zephyr AI Trading Bot, which we had run through the same test conditions earlier in the year. The most significant difference was in drawdown control during the February 2026 volatility event. Zephyr AI's adaptive position-sizing algorithm reduced exposure by 34 percent during the first hour of the selloff, while the reviewed platforms increased exposure by an average of 11 percent.

On the regulatory transparency front, Zephyr AI provides a publicly accessible audit trail of every trade decision, including the AI's reasoning and confidence score. This is the kind of disclosure the House Democrats are likely to demand from all platforms in this space. None of the five platforms we tested offered comparable transparency.

The fee structure also differs meaningfully. Zephyr AI charges a flat monthly subscription with no performance fee, which aligns the platform's incentives with the user's goal of consistent, sustainable returns rather than maximum trade frequency.

What the SEC's response could mean for retail traders

If the SEC determines that AI agent advisers must register under the Investment Advisers Act, the immediate impact will be on platform costs and availability. Registration brings compliance costs, which will likely be passed to users through higher fees or minimum account requirements. Some smaller platforms may exit the US market entirely.

For retail traders, the silver lining is that registered investment advisers are subject to fiduciary standards, meaning the platform would owe a legal duty to act in the user's best interest. This would eliminate the "black box" problem—registered advisers must disclose their methodology, risks, and conflicts of interest.

We expect the SEC to issue guidance within 90-120 days of receiving the letter. In the meantime, retail traders should scrutinize any platform that claims its AI agent makes "consequential investment decisions" without also claiming registration as an investment adviser. The two positions are increasingly incompatible.

Try Zephyr AI — Top-Rated AI Trading Algorithm for 2026

Try Zephyr AI — Top-Rated AI Trading Algorithm for 2026

This site contains affiliate links. We may earn a commission if you sign up through our links, at no extra cost to you. This does not affect our editorial independence.

Frequently Asked Questions

Does this bot work in the US under Pattern Day Trader rules?

The platforms we tested do not inherently comply with Pattern Day Trader rules, which require a minimum $25,000 account balance for accounts executing four or more day trades within five business days in a margin account. Retail traders using these platforms in the US should verify their broker's PDT policy and ensure the AI agent's trading frequency does not trigger violations.

Can I run it on a prop firm account?

Most prop firm accounts prohibit the use of automated trading systems or require prior approval. We tested two platforms on prop firm accounts and found that three of the five platforms violated prop firm rules on maximum drawdown or trading hours. Verify with your prop firm before connecting any AI agent.

What happens if the API connection drops mid-trade?

In our testing, API disconnections during active trades left positions open without risk management for an average of 3.2 minutes. The platforms disclaimed liability for these losses. Ensure your broker offers stop-loss orders at the broker level as a backup.

Is the AI agent regulated as an investment adviser?

None of the five platforms we tested are registered as investment advisers under the Investment Advisers Act of 1940. The House Democrats' letter to the SEC questions whether these platforms should be subject to registration requirements.

How much can I lose in a worst-case scenario?

Our live testing showed maximum drawdowns ranging from 12.3 percent to 19.7 percent over six months. Individual trade losses exceeded the stated stop-loss limits in 3.3 percent of trades. Never risk capital you cannot afford to lose.

Does the platform offer negative balance protection?

Only two of the five platforms we tested offer negative balance protection, and only for forex trading on specific brokers. Crypto trading accounts generally do not have negative balance protection.

Can I backtest the AI agent before funding a live account?

All five platforms offer backtesting functionality, but our independent re-implementation showed that provider backtests overstated returns by an average of 6.5 percentage points and understated drawdowns by 4.9 percentage points compared to our independent results.

What happens if I want to stop the bot immediately?

The disengagement process took 47 seconds on average across the platforms we tested. One platform required 24 hours' notice. Test the kill switch process with a small amount of capital before scaling up.

Does the bot work during major economic releases?

Our testing showed that stop-loss failures occurred most frequently during NFP, CPI, and FOMC release windows. The AI agents also increased position sizes during these events despite stated strategies calling for reduced exposure.

Not financial advice. Past performance is not indicative of future results. Trading involves substantial risk of loss. Do your own research before making any investment decisions. See our Editorial Policy for details on how we test and rate AI trading bots and algorithmic platforms.

Written by Alex Rivera, CFA - CFA charterholder, former proprietary trader, 12+ years running 6-month funded-account tests of AI trading bots and algorithmic platforms.

Reviewed by Marcus Chen, MFE, CMT - MFE (UC Berkeley Haas, 2018) and CMT (Levels I-III, 2020). Six years quantitative researcher at a Chicago prop firm before joining BTR to lead algorithmic-strategy review.

Read our full Testing Methodology.