Just now, global AI giant Anthropic officially issued an ultimatum: starting at 3:00 PM Eastern Time on April 4, OpenClaw will be forcibly removed from the whitelist for Claude subscription services. Developers worldwide were stunned.

Author and source: AI World

Just now, an unprecedented "isolationist" event occurred in the AI community.

Anthropic has officially banned the use of its own plans to access OpenClaw!!!

Boris Cherny, the creator of Claude Code, announced:

Starting at 3:00 PM Eastern Time on April 4 (3:00 AM Beijing Time on April 5), Claude will block all third-party tools, and these tools can only be accessed through additional plans or the API.

This means that thousands of developers and startup teams relying on OpenClaw for improved efficiency suddenly lost their access to unlimited usage, forcing them to switch to an extremely expensive pay-as-you-go model.

A loyal Claude user on OpenClaw took a hard hit.

The timing of this announcement is also highly intriguing—Peter Steinberger, the founder of OpenClaw, recently joined OpenAI, and Anthropic’s intentions are now clear!

This can be described as a business vendetta disguised as policy.

This message quickly rose to the top of the developer community Hacker News.

Let’s remember this day: April 4, 2026, the day the AI industry shifted from open collaboration to dominance by major players.

Claude's official email proves that, this time, Anthropic is targeting OpenClaw!

This policy will be enforced starting April 4 on OpenClaw first, but applies to all third-party toolchains and will soon be extended to additional tools.

As a goodwill gesture, Anthropic has provided a one-time credit equal to one month's subscription fee, valid until April 17.

The hammer falls

Note that the timing of this email from Anthropic is highly inappropriate.

On Friday evening, April 3rd, the favorite time for internet companies to release bad news.

Anthropic has notified thousands of OpenClaw users: Starting tomorrow, your Claude subscription quota can no longer be used for OpenClaw. If you wish to continue, you will be charged on a pay-as-you-go basis.

https://x.com/VadimStrizheus/status/2040199979927482618

After three months of crackdown, the final blow has finally fallen.

Catalyst: A Defection Powerful Enough to Reshape the Landscape

Why is Anthropic, at this juncture, ruthlessly targeting an open-source tool with such disregard for propriety?

Because Peter Steinberger, the visionary behind OpenClaw and the "father of lobster," has joined their archrival, OpenAI.

Once, Peter Steinberger was one of the most knowledgeable developers in the Claude ecosystem, and his OpenClaw made Claude incredibly user-friendly.

Now, for Anthropic, OpenClaw has become a Trojan horse from the enemy camp.

Anthropic believes OpenClaw is no longer just an efficiency tool, but a "data collector" encroaching on its own backyard.

Since the founder is now with OpenAI, you can no longer use my subscription quota.

Peter himself reluctantly spoke out, implying that Anthropic is “shutting the door and beating the dog” and “freeloading” on the open-source community:

I, along with Dave Morin (a board member of OpenClaw), attempted to persuade Anthropic to remain calm.

But in the end, all we could achieve was to delay the arrival of this day by one week.

https://x.com/steipete/status/2040209434019082522

Developers are in distress; budgets have skyrocketed overnight.

For ordinary developers, this ban is essentially a game-changing blow.

Previously, many developers achieved extremely cost-effective automated workflows by subscribing to Claude's fixed monthly plan and combining it with OpenClaw's powerful API.

Spending $20 on a Claude Pro plan allows the lobster to have Claude working 24/7. The same usage via API could result in a bill running into thousands of dollars.

One is a Max subscription capped at $200, and the other is a four-digit API fee.

Now, this path has been cut off by Anthropic itself.

Pay-as-you-go means abandoning monthly subscription plans, making costs highly unpredictable. Many small and medium-sized teams had their AI budgets locked in on a monthly basis, but now they could face sudden cost spikes at any time.

Even more harshly, if you don’t want to pay this exorbitant toll fee, you must endure the pain of completely rebuilding your business logic within 24 hours.

One word—amazing!

The lobster stirs the江湖, and grievances have long begun.

The feud between the Father of Lobster and Anthropic began long ago.

Steinberger publicly complained that dealing with Anthropic was "basically all through legal letters."

First tactic: Brand separation.

At the end of January, a cease-and-desist letter forced Clawdbot to change its name.

Second tactic: Technical blocking.

On January 9, Anthropic quietly implemented a server-side detection: any subscription token not originating from the official Claude Code client is immediately rejected.

The core gameplay of OpenClaw: lose everything overnight.

Third tactic: Qualifying the terms.

In mid-February, the Terms of Service were updated: Using OAuth tokens for Free, Pro, or Max accounts in any third-party tools constitutes a violation.

The fourth and most powerful strategy is to bottom-feed on features.

Claude Cowork has launched Dispatch, enabling remote control of the desktop Claude app from your phone; Claude Code has introduced Channels, integrating with Telegram and Discord.

Within four weeks, OpenClaw's core features were officially replicated one-to-one!

In the words of AI influencer Matthew Berman: "They just built OpenClaw themselves."

Tech media Semafor reported this morning that when Anthropic’s Chief Business Officer, Paul Smith, was asked whether clients were requesting the company to build its own OpenClaw, he acknowledged that this is indeed the case.

https://www.semafor.com/article/04/03/2026/anthropic-eyes-its-own-version-of-openclaw

And today, the email on April 4th was the final blow.

Open strategy: Force-promoting our own "favorite child," Claude Cowork

While banning OpenClaw, Anthropic also revealed its true intentions.

They began hinting strongly: Stop using those unreliable third parties—try our native Claude Cowork today!

Claude Cowork enables deeper control over the coding environment and computer interface.

Actually, this isn't exclusive to Anthropic either.

This is precisely the "platform lock-in" strategy that giants excel at:

Step 1: Attract developers using third-party open-source tools to expand the ecosystem.

Step 2: Find excuses (such as security concerns or infrastructure strain) to ban third parties.

Step 3: Force users to migrate to the platform’s own native integrated tools, which are more expensive and strictly controlled.

This typical "vertical integration" approach allows them to maintain tight control over access and user experience, while gradually diminishing the advantages of those "uncontrolled connectors" within the ecosystem.

Interestingly, OpenAI chose an entirely opposite path around the same time.

OpenAI explicitly permits Codex subscriptions for use in third-party clients such as OpenClaw.

In March, it went a step further by announcing free ChatGPT Pro access for open-source project maintainers, with OpenClaw explicitly named on the beneficiary list. The people Anthropic wanted to push out, OpenAI welcomed with open arms!

No one is entirely innocent.

Fairly speaking, Anthropic has its point.

Every user subscribing for $200 who generates API usage in the thousands of dollars is causing the company to lose money.

Third-party tools that bypass official telemetry, spoof client identities, and create monitoring blind spots are real engineering and security risks.

Moreover, Steinberger is now sitting in OpenAI’s office—OpenClaw has, in Anthropic’s eyes, shifted from being a “slightly annoying freeloader” to a “spy sent by the other side.”

Moreover, OpenClaw itself is not clean.

A critical vulnerability with a CVSS score of 8.8 (CVE-2026-25253) has been disclosed, allowing attackers to steal users' authentication tokens via a malicious link.

https://nvd.nist.gov/vuln/detail/CVE-2026-25253

On the public internet, security agencies have identified over 30,000 openly accessible OpenClaw instances.

But these justifications don't eliminate the developers' feeling of being cheated.

The platform first opens the door, waits for people to build the structure, and then announces a different plan—this script has been played out too many times.

X shutting down third-party clients, Apple tightening App Store policies, and Google eliminating free APIs—all are the same story.

Every time, it's the developers who get hurt.

The open ecosystem enters its twilight as AI giants shut themselves off.

Today, Anthropic's most devastating move has made us realize a heartbreaking fact—

The era of free, open AI for developers is coming to an end.

Once, we thought AI would evolve like the early internet—based on shared protocols and community-driven development. But in reality, model owners hold all the power and can instantly render the efforts of thousands obsolete.

The countdown to the deadline of 3:00 PM on April 4 has begun.

Developers face three harsh choices—

Should you swallow the bitter pill, endure exorbitant usage-based fees, and continue using OpenClaw until your budget runs out?

Or fully migrate and submit to Anthropic’s native tools, accepting platform lock-in?

Or walk away in anger and completely abandon Claude.

Finally, while the giants are busy clashing like gods, don’t forget that it’s tens of thousands of developers, line by line of code, who built the throne upon which your valuations of hundreds of billions rest.

This time, Anthropic won the competition but lost the community's trust completely!

Reference materials:

https://www.techbuzz.ai/articles/anthropic-blocks-openclaw-from-claude-subscriptions

https://www.theverge.com/ai-artificial-intelligence/907074/anthropic-openclaw-claude-subscription-ban