Anthropic has accused Alibaba of attempting to extract capabilities from its Claude AI system through what it describes as a large-scale and coordinated pattern of automated usage. The claim was outlined in a letter submitted to US Senate Banking Committee members and White House officials.
Anthropic alleged, according to Bloomberg, that Alibaba Group Holding Ltd. waged a large-scale effort to “illicitly” access its Claude artificial intelligence model using thousands of fraudulent accounts, bypassing the developer’s decision to restrict its products in China.
The company has framed the activity as a potential attempt to replicate internal model behaviour using repeated querying patterns rather than standard commercial use.
Tens of millions of prompts flagged in suspected campaign
According to Anthropic, the activity occurred between April 22 and June 5, 2026, and involved approximately 28.8 million interactions with Claude. These interactions were reportedly distributed across roughly 25,000 access accounts, which Anthropic describes as automated or coordinated identities designed to spread usage and avoid detection limits.
The company alleges the traffic pattern was consistent with systematic probing of advanced capabilities, including coding performance, reasoning behaviour, and tool-use functions embedded in the model.
Anthropic reportedly believes the campaign was linked to operators associated with Alibaba’s Qwen AI lab, and that it targeted Claude’s most prized capabilities, including software engineering and agentic reasoning.
Distillation concerns sit at the core of dispute
Anthropic’s complaint centres on the concept of “model distillation,” a method in which outputs from a large AI system are used to train another model in order to reproduce aspects of its performance.
The company argues that while distillation can be legitimate in controlled internal optimisation, large-scale extraction of outputs from a competitor’s frontier model may represent improper appropriation of proprietary capabilities.
Anthropic has previously warned that such techniques could reduce the competitive advantage of frontier AI developers by enabling faster replication of costly model behaviours.
Alibaba (BABA) drops 3 percent amid AI allegations
Alibaba Group Holding Ltd. (BABA) stock has been in a sustained intraday downtrend, extending a sequence of lower highs and lower lows that reflects persistent bearish momentum rather than a single event-driven shock.
Following the news of Anthropic’s allegations, the stock continued to drift lower, with sentiment already weak and selling pressure remaining in place as the headline added to the downside momentum.
RSI (14) is in the mid-20s, showing deeply oversold conditions. While this points to stretched downside momentum and possible short-term stabilization, there is no reversal signal yet and selling pressure remains dominant.
Overall, the trend structure remains decisively negative, with price action still controlled by sellers. A clear break above prior intraday resistance around 104 to 106 dollars would be required to shift the current downside outlook.
