The line between genuine robotic independence and hidden human control has become the central controversy in the humanoid robotics race. Figure AI CEO Brett Adcock is defending his company's flagship Figure 03 robot against claims that it relies on teleoperation, a technique where humans remotely guide the machine's actions rather than letting it think and act on its own .

What Exactly Is the "Smoking Gun" Evidence Against Figure?

The controversy erupted after Figure's appearance on the Shawn Ryan Show, where Adcock demonstrated the Figure 03's capabilities in an outdoor setting. The episode was meant to showcase the Helix 02 architecture, a "Software 2.0" approach that replaces hand-coded programming with end-to-end neural networks, allowing the robot to learn directly from visual data .

However, skeptics quickly identified what they saw as a critical flaw. Nima Zeighami, co-founder of the humanoid robot fighting company REK, posted on X that he spotted the robot beginning to turn around before Adcock gave the verbal command to do so. This timing gap suggested the robot might be receiving instructions from a human operator behind the scenes, not making autonomous decisions .

"This video is the smoking gun that Figure's robots are teleop'd and not autonomous. Notice how the robot starts turning around before Brett says 'alright turn around'," stated Nima Zeighami.

Nima Zeighami, Co-founder at REK

The accusation cuts deep because Adcock has been perhaps the industry's most vocal critic of human-in-the-loop strategies. He has famously dismissed teleoperation as "soy stuff" and positioned Figure's AI-first approach as fundamentally different from competitors like 1X, which use supervised "Expert Mode" where humans actively guide robots through tasks .

How Did Figure Respond to the Autonomy Challenge?

When the skepticism followed Figure to its most prestigious stage yet, the White House, Adcock doubled down on his claims. After posting photos of the Figure 03 walking alongside First Lady Melania Trump during the "Fostering the Future Together" summit, followers questioned whether the "historic" moment was truly autonomous or if someone was "behind the curtain" controlling the robot .

Adcock offered a definitive rebuttal to these questions. When asked directly by a user named CaffeineMachine, he responded with absolute certainty about the robot's independence during the White House appearance .

"Autonomous. No human was in the loop for this," declared Brett Adcock.

Brett Adcock, CEO at Figure AI

At the White House event, the Figure 03 utilized integrated speech and reasoning capabilities to deliver a multilingual welcome to attendees, demonstrating what Figure claims is genuine autonomous decision-making in a high-stakes public setting .

Why This Debate Matters for the Entire Industry

The autonomy question represents far more than a technical disagreement between rival companies. It goes to the heart of what makes a humanoid robot valuable. A truly autonomous robot that can navigate complex environments and make real-time decisions without human guidance would represent a genuine breakthrough. A robot that requires constant human oversight, by contrast, is essentially an expensive remote-controlled toy .

Figure's current strategy relies on the claim that its robots compute movement instructions directly from visual data, bypassing the rigid motion planning systems used by earlier robot generations. If Figure is indeed achieving what it calls "room-scale" autonomy without human intervention, it would represent a significant lead over competitors in the field .

The stakes are extraordinarily high. Figure is valued at $39 billion, and the company is scaling production toward 50,000 units at its BotQ facility. As production ramps up, the pressure to demonstrate reliability in genuinely unpredictable and unscripted environments will only intensify .

How to Evaluate Robot Autonomy Claims in the Future

• Timing Analysis: Watch for whether robots begin executing commands before receiving verbal instructions, which could indicate pre-programmed sequences or hidden human control rather than real-time autonomous decision-making.
• Environmental Variability: Assess whether robots perform consistently in genuinely novel environments they have never encountered before, rather than in controlled settings or locations where they may have been pre-trained.
• Real-Time Adaptation: Evaluate whether robots can respond to unexpected obstacles, changes in instructions, or environmental surprises without pausing or reverting to scripted behaviors that suggest human intervention.
• Independent Verification: Look for third-party testing and transparent documentation of how robots make decisions, rather than relying solely on company-produced demonstrations.

The autonomy versus teleoperation debate will likely define the humanoid robotics industry for the next several years. As companies race to prove their robots can think and act independently, the technical details of how those claims are verified will become increasingly important to investors, regulators, and the public .

For now, the "smoking gun" remains a matter of interpretation. But the stakes for Adcock's $39 billion startup and the broader credibility of the humanoid robotics sector could not be higher. The company that can definitively prove genuine autonomy in real-world conditions may ultimately win the race to make humanoid robots a practical reality in factories, warehouses, and beyond.