Unbiased AI-powered news
Alibaba directed employees to stop using Anthropic's Claude Code after the tool flagged connections from China. The company instructed staff to switch to its internal Qoder platform instead.
thenextweb.comAlibaba has ordered its employees to stop using Anthropic's Claude Code coding tool. The directive followed detection that the tool inspected user timezone and proxy information to flag connections originating from China and inserted subtle markers. Employees were told to use Alibaba's internal Qoder platform as the replacement.
The ban was first reported by Chinese media. Anthropic stated that Alibaba had deployed around 25,000 fake accounts to train its own models on Claude outputs. The company described the activity as a distillation strike in a letter sent to two U.S.
Senators. Anthropic employee said on X that the flagging feature was an experiment launched in March to curb abuse and distillation attempts.
Mark Zuckerberg told employees Thursday that development of AI agent technology has fallen behind internal targets. The company also paused a mandatory employee monitoring program last month after a leak and cut 10 percent of its workforce in May.
TechCrunchMeta released Pocket, an app that lets users generate and share interactive mini games through text prompts. The app first appeared on the App Store and Google Play on June 29, 2026, though it remained unavailable for download in the United States as of July 2.
theconversation.comManagers at AI startups direct engineers to use different models depending on task difficulty. Coinbase CEO Brian Armstrong projected that 80 percent of workloads will shift to cheaper models within 12 to 18 months. Model router adoption among firms rose from 1 percent last year…