Google Faces Backlash Over “Draconian” AI Ban as OpenClaw Conflict Escalates

New Delhi, February 23, 2026: A heated dispute has erupted between Google and the open-source community following a massive wave of account bans targeting users of OpenClaw, a popular AI agent framework. The conflict reached a boiling point on Monday, February 23, 2026, when OpenClaw creator Peter Steinberger publicly condemned Google’s tactics as “draconian,” while the tech giant doubled down on its security-first stance.

The crackdown has sent shockwaves through the developer community, particularly because it has targeted high-paying subscribers of Google’s AI Ultra tier, many of whom pay upwards of $250 per month for premium access to Gemini models.

The “Draconian” Crackdown: A Lack of Warning

The controversy began when developers started receiving “Error 403” messages, indicating that their access to Google Antigravity—the backend infrastructure powering Google’s latest coding AI—had been terminated. Unlike typical service interruptions, these bans were reportedly issued without prior notification or a grace period.

Taking to social media, Peter Steinberger, who was recently recruited by OpenAI to lead its personal agents division, expressed his frustration with Google’s lack of communication.

“Pretty draconian from Google. Be careful out there if you use Antigravity,” Steinberger posted. “Even Anthropic pings me and is nice about issues. Google just bans?”

Steinberger hinted that he may be forced to remove Google support from OpenClaw entirely to protect users from losing their primary Google accounts, which are often tied to essential services like Gmail, YouTube, and Workspace.

Google’s Defense: Fighting “Malicious Usage”

Google has been quick to defend its actions, citing a necessity to maintain service integrity. Varun Mohan, lead for Google Antigravity, stated that the company observed a “massive increase in malicious usage” coming from third-party wrappers that bypass official channels.

The Core Issues:

• OAuth Misuse: Google alleges that OpenClaw users were utilizing unauthorized OAuth tokens to access Antigravity. These tokens are intended for first-party Google applications, and using them in third-party tools is a direct violation of the Terms of Service.
• Service Degradation: Google claims that automated loops and high-frequency calls generated by AI agents were degrading the quality of service for legitimate users.
• Economic Arbitrage: By using OpenClaw to route consumer-tier subscriptions (like AI Ultra) into professional developer workflows, users were effectively bypassing the more expensive per-token pricing of the official Gemini API.

Security Under the Microscope

Beyond the policy disputes, the safety of the OpenClaw framework itself has come into question. Just weeks ago, security researchers identified CVE-2026-25253, a critical “1-click” remote code execution vulnerability in OpenClaw. This flaw allowed attackers to steal authentication tokens and take over a user’s local machine via a single malicious link.

While Steinberger and the OpenClaw foundation have since patched the vulnerability, the incident gave corporate giants like Meta and Google additional leverage to justify banning the tool from their networks, citing it as “potentially unwanted application” (PUA) or even malware.

The Broader “AI Turf War”

The clash reflects a larger shift in the industry as tech giants move to “close the gates” around their most powerful models.

Company	Recent Action on Third-Party Agents
Google	Mass bans on Antigravity users; strict OAuth enforcement.
Anthropic	Updated legal terms to explicitly ban third-party OAuth usage for Claude.
OpenAI	Acquired the OpenClaw creator and remains the only major lab to whitelist the tool.

For the average developer, the message is clear: using unofficial “wrappers” to access premium AI models is becoming a high-stakes gamble. With Google accounts now acting as the “digital passport” for many individuals’ lives, a single ban in the AI lab can result in a total digital lockout.

What’s Next for Developers?

As the “AI agent wars” continue, developers are being urged to migrate away from OAuth-based workarounds and toward official API keys. While this often comes with a higher price tag, it offers the only guaranteed protection against the “ban hammer” currently swinging through Silicon Valley.