Google caused controversy among some developers after restricting the use of its new AntiGravity this weekend and today, Monday, February 23. "vibe coding" platform, charge "Malicious use."
SSome users who were using the open source autonomous AI agent OpenClave in conjunction with agents built on AntiGravity, as well as those who had OpenClave agents linked to their Gmail, claimed on social media that they had lost access to their Google accounts.
According to Google, users were using AntiGravity to access a large number of Gemini tokens through third-party platforms such as OpenClaw, which impacted systems for other AntiGravity customers.
The move has alienated many users, underscoring the architectural and trust issues that have arisen with OpenGL. The time of Google’s action is specifically indicated. Just a week earlier, on February 15, OpenAI CEO Sam Altman announced that OpenAI creator Peter Steinberger had joined OpenAI to lead its “next generation of personal agents.” While OpenClaw remains an open-source project under an independent foundation, it is now financially supported and strategically directed by Google’s primary rival.
By cutting off OpenGL’s access to AntiGravity, Google is not only protecting its server load; This is effectively breaking a pipeline that allows OpenAI-adjacent tools to take advantage of Google’s most advanced Gemini models.
Varun Mohan, a Google DeepMind engineer and former CEO and founder of Windsurf, said in an X post that the company saw “malicious use” that led to the service degradation.
“We are seeing a huge increase in malicious use of the AntiGravity backend, which has significantly degraded the quality of service for our users. We needed to find a way to immediately shut off access to users who are not using the product as intended. We understand that a subset of these users were not aware that it was against our TOS. [Terms of Service] And they will find a way back but our capacity is limited and we want to be fair to our end users,” the post said.
A spokesperson for Google DeepMind told VentureBeat that the move is not meant to permanently ban the use of AntiGravity to access third-party platforms, but rather to align its use with the platforms’ terms of service.
Unsurprisingly, Google’s move has caused an uproar among OpenClaw users, including OpenClaw creator Peter Steinberger, who announced that OpenClaw would be removing Google support as a result.
Infrastructure and connection uncertainty
OpenClaw emerged as a way for individual users to run shell commands and access local files, fulfilling a key promise of AI agents: running workflows efficiently for users.
But, as VentureBeat has repeatedly reported, this can often lead to security and guardrails issues. There are companies that are creating ways for enterprise customers to access OpenClaw securely and with a governance layer, although OpenClaw is so new that we should expect more announcements soon.
However, Google’s move was not seen as a security issue, but rather one of access and runtime, suggesting there is still significant uncertainty when users want to bring something like OpenClaw into their workflow.
This is not the first time that developers and power users of agentic AI have had their access cut. Last year, Anthropic blocked access to the cloud code because the company claimed some users were abusing the system by running it 24/7.
This highlights the disconnect between companies like Google and OpenClaw users. OpenClaw offers many interesting possibilities for creating workflows with agents. However, because it is constantly evolving, users may inadvertently violate TOS or rate limits.
Mohan said Google is working to bring back banned users, but whether that means the company will amend its ToS or explore a secure connection between OpenClaw agents and antigravity models remains to be seen.
For developers, the message is clear: the era of "bring your own agent" A marginal model is coming to an end. Providers are now prioritizing vertically integrated experiences where they can capture 100% of telemetry and subscription revenue, often at the expense of the open-source interoperability that defined the early days of the LLM boom.
affected users
Several users said on both the Y Combinator chat board and X that they no longer had access to their Google accounts after running OpenGL instances for certain Google products.
Google’s move reflects broader industry changes "walled garden" Agent ecosystem. Earlier this year, Anthropic introduced "customer fingerprinting" It effectively locks out third-party wrappers like OpenClause to ensure that its cloud code environment remains the exclusive interface to its models. For developers, the message is clear: the era of "bring your own agent" A marginal model is coming to an end. Providers are now prioritizing vertically integrated experiences where they can capture 100% of telemetry and subscription revenue, often at the expense of the open-source interoperability that defined the early days of the LLM boom.
Some people have said that they will no longer use Google or Gemini for their projects. Right now, those who still want to continue using AntiGravity will have to wait until Google finds a way for them to use OpenClaw and access the Gemini token in a way that Google deems “appropriate.”
Google DeepMind reiterated that it has only cut off access to AntiGravity, not other Google applications.
Conclusion: Enterprise Takeaways
For enterprise technical decision makers, "anti gravity restraints" The agent serves as a definitive case study in the risks of dependency. As the industry moves from chatbots to autonomous agents, the following realities must now dictate strategy:
- Platform fragility is the new normal: $250/month lockout "extremely" Users prove that even high-paying enterprise customers have little to gain when a provider decides to change it "fair use" Definitions. Relying on OAuth-based third-party wrappers for core business logic is now a high-risk gamble.
-
Rise of local-first governance: With OpenClaw moving to an OpenAI-enabled foundation and Google/Anthropic consolidating their cloud, enterprises should prioritize agent frameworks that can run "local-first" Or within a VPC. "token loophole" That OpenClaw exploit is being shut down; Agentic scale in the future will require direct, high-cost API contracts rather than subsidized consumer seats.
-
Account portability as required: The fact that the user "Lost access to their Google accounts" Underscores the danger of bundling development environments with primary identity providers. As far as possible, decision makers should separate AI development from the core corporate identity (SSO) to avoid a single TOS violation crippling the entire team’s communications.
Ultimately, it marks the end of the antigravity phenomenon. "wild West" For AI agents. As Google and OpenAI make their claims, enterprises will have to choose between the stability of walled gardens or the complexity (and cost) of truly independent, self-hosted infrastructure.
<a href