Anthropic officially "banned" OpenAI. GPT-5 is about to be released, but it was revealed that it was developed using Claude Code?

Anthropic Blocks OpenAI's Claude API, Alleging Violation of Terms by Using Claude to Support GPT-5 Development and Safety Testing. This Move Marks a New Round of Blockade Among AI Giants Regarding Data and Interfaces, with API Becoming a Strategic Resource for Market Access and Innovation, Sparking Heated Discussion and Regulatory Attention. Flipping the Table, Unplugging the Network, Completely Blocking! This Tuesday, Two AI Giants Tore Apart Their Relationship. According to Multiple Informed Sources, Anthropic Has Cut Off OpenAI's API Access to All of Its Large Language Models. In Other Words, OpenAI Has Been Completely "Blocked" by Its Old Rival Anthropic. Anthropic's Reason Was Simple and Brutal: OpenAI Openly Violated the Agreement! Anthropic's Spokesperson Christopher Nulty Issued a Highly Charged Statement: Claude Code (Anthropic's Programming Tool) Is Now an Essential Weapon for Programmers Worldwide. Therefore, We Were Not Surprised to Learn That OpenAI's Own Technical Personnel Were Using Our Programming Tools Before the GPT-5 Release. "Unfortunately, This Behavior Seriously Violates Our Service Terms." According to Anthropic's Commercial Terms, No Customer May Use Its Services to "Build Competing Products or Services, Including Training Competitive AI Models," Let Alone "Reverse Engineer or Replicate" Its Services. The Background of This Blocking Incident Comes at a Crucial Moment - It Is Rumored That OpenAI Is About to Release the Revolutionary New Model GPT-5, Whose Core Selling Point Is Its Extremely Strong Programming Ability. So, Just as OpenAI Is Preparing to Launch the Rumored GPT-5 (Said to Have Stronger Coding Capabilities and Safety Features), Anthropic's API Control Strategy Seems to Be a Proactive Defense Involving Competition and Safety Positioning. For OpenAI, Wanting to Overturn the World with a New Model While "Stealing" Techniques from Its Strongest Competitor? This Drama Is Too Explosive! So, What Exactly Was OpenAI Doing with Claude? It Was Revealed That OpenAI Was Not Simply Using the API to "Chat" Like Ordinary Users. Instead, They Deeply Integrated Claude into Their Internal Tools Using Special Developer Permissions (API). Their Purposes Were Mainly Two-Fold: Probing the Opponent: Systematically Assessing Claude's Real Capabilities in Programming, Creative Writing, and Other Areas, Conducting Comprehensive Benchmarking Against Their Own AI Model. Security "Red and Blue Team Confrontation": Testing Claude's "Bottom Line" with Various Tricky and Dangerous Prompts, Such as Topics Involving CSAM, Self-Harm, Defamation, etc., Observing Its Responses to Help OpenAI Evaluate and Adjust Its Own Model's Safety Guardrails. Faced with the Sudden Blockade, OpenAI Appeared Both Shocked and Disappointed. OpenAI's Chief Communications Officer Hannah Wong Responded: Evaluating Other AI Systems to Measure One's Own Progress and Improve Safety Levels Is Already an Industry Practice. We Respect Anthropic's Decision, But Honestly, This Is Disappointing. Don't Forget, Our API Has Always Been Open to Them. This Response Implied: We Are Open to You, But You Close the Door on Us? That's Unfair! Interestingly, Anthropic's Spokesperson Nulty Subsequently Stated That the Company Would "Continue to Ensure OpenAI Has API Access for Benchmarking and Safety Assessment" Because This Is an "Industry Standard". However, This Statement Contradicts the Current Blocking Action, and Anthropic Has Not Provided Further Explanation. In the Tech World, Using API Interfaces to Strangle Competitors Has Long Been a Regular Weapon. Facebook Once Did the Same to Twitter's Vine Short Video App, Triggering an Antitrust Investigation. Just Last Month, Salesforce Also Restricted Competitors' Access to Its Data via Slack API. This Isn't Even Anthropic's First Time "Unplugging the Network". Last Month, When Rumors Spread That OpenAI Might Acquire AI Programming Newcomer Windsurf, Anthropic Quickly Cut Off Windsurf's Direct Access to Claude. Anthropic Said They Only Wish to Support "Long-Term Partners". At the Time, Anthropic's Chief Scientist Jared Kaplan Spoke Candidly to the Media: "Selling Our Claude to OpenAI? The Thought Itself Seems Strange." Some Suggest That Open Source Is the Only Solution. Some Even Joked, "GPT-5, Produced by Claude Code". Just a Day Before This Blockade of OpenAI, Anthropic Had Just Announced New Usage Frequency Restrictions on Claude Code Due to "Explosive User Growth" and Some Users Violating Service Terms. Anthropic's "Unplugging" Is Not Just a "Tactical Strike" Against OpenAI, But Reflects the AI Competition Entering a Phase of "Data and Interface Blockade". For Developers, Startups, and Even Regulatory Agencies, API Is No Longer Just a Technical Detail, But a Strategic Resource Concerning Market Access and Innovation Freedom. Everything Was Predictable - The War for the AI Throne Has Completely Heated Up!