阿里Qoder开源AI编程体检工具，Claude Code和Codex也能用
Related
Meta restricts employee use of Claude and Codex to prevent "model distillation" violations.
According to Beating's monitoring, Meta is strictly limiting its applied AI engineering team's use of Anthropic's Claude Code and OpenAI's Codex, and has even suspended some related tasks to strengthen compliance reviews. Meta is concerned that employees' over-reliance on external tools to develop internal alternatives could lead to the output of competing models seeping into its own training data. This "model distillation" practice could violate the terms of service of competing products and cause serious conflicts with partners. Currently, Meta only allows employees to use external AI for routine tasks such as setting up workflows, cleaning code, and building test infrastructure, and the output must be strictly reviewed by humans; however, it strictly prohibits the use of external models to generate programming challenges to evaluate its internal models, and also prohibits the use of external AI to find code vulnerabilities or devise test tasks. This move is also part of Meta's efforts to reshape its internal toolchain and control its multi-billion dollar internal AI operating costs by promoting its self-developed coding assistant MetaCode (formerly DevMate).
Anthropic admitted that Claude Code had embedded steganography code targeting Chinese users, calling it an "abuse prevention experiment," and promised to roll back the code tomorrow.
According to Beating's monitoring, Thariq, an engineer on Anthropic's Claude Code team, publicly responded to the recent controversial "spy code" leak. He admitted that in March of this year, an experimental mechanism was embedded in the product. This mechanism detected whether the system timezone was Asia/Shanghai or Asia/Urumqi, whether the proxy hostname matched a list of Chinese resellers, and the keyword "AI Lab," and used special punctuation marks to inject hidden marker information into system prompts in a steganographic manner. He stated that the mechanism was intended to "prevent unauthorized resellers from abusing accounts and model distillation," but emphasized that the team has since implemented stronger protective measures and "has always intended to take it offline." The relevant PR has been merged, and it is expected to be completely rolled back in tomorrow's version release. This leak was made public on June 30 by the security account @IntCyberDigest, accompanied by two screenshots of code showing that Claude Code performed environmental fingerprinting on Chinese users without their knowledge. While Thariq's response was a direct admission, the timeline of "launching in March and only accelerating its withdrawal after being exposed" has still sparked widespread skepticism within the community. The comments section almost unanimously criticized Anthropic for "only announcing its withdrawal after being caught" and "secretly monitoring users without notifying them," severely damaging the company's long-standing image of "prioritizing security and ethics." --------------------------------- Click the original link below to join the Beating · Lark AI news channel for 24/7 monitoring of global AI hot topics and news.
The culprit behind the exhausted Codex credit limit has been identified; OpenAI has fixed multiple vulnerabilities and implemented a third round of full-scale compensation resets.
According to Beating's monitoring, the cause of the abnormal consumption of credit limits in Codex, a programming intelligence agent under OpenAI, has been officially identified. Tibo Sottiaux, the core product manager, announced that the team has deployed a patch to fix the issue across all users. In addition to resetting the credit limits for all users, all users will also receive an additional reset card valid for 24 hours. The excessive consumption was not caused by a single vulnerability, but by a combination of multiple minor backend issues and display false alarms. At the operational level, the system's automatic review was too frequent, unexpectedly triggering too many sub-agent tasks, and the backend suggestion function repeatedly ran and retried after failure, consuming tokens exponentially. At the display level, automatic review was incorrectly classified as GPT-5.4 consumption, and failed or rate-limited requests were also incorrectly displayed as credit consumption in the frontend charts, directly causing a credit shortage for all users. Currently, the official team has deployed a hotfix patch simultaneously on the billing backend, desktop client, and CLI terminal. In addition to resetting the credit limits for all users, only successful interaction requests will be recorded in the Turn statistics chart in the future. Although the erroneous data in the historical charts cannot be changed, the actual token consumption after the update will be significantly reduced.
OpenAI Codex's anti-abuse mechanism misjudged and caused a surge in credit limits, prompting the official team to reset the credit limits for all users.
According to Beating's monitoring, Tibo Sottiaux, core product lead at OpenAI, stated that due to a misjudgment by the system's built-in anti-fraud and anti-abuse mechanisms leading to accelerated depletion of user credit limits, the official team has urgently reset the usage limits for all Codex users free of charge. The abnormal credit limit depletion issue actually began around June 16th and fully erupted on social media platforms on June 25th. Hundreds of developers reported on GitHub and social media that Codex token consumption was abnormally high, with some accounts experiencing a 10 to 20 times faster depletion rate than usual. A paid user with a $200/month subscription reported that their additional $40 temporary credit limit was exhausted within 3 to 4 hours. Other developers reported that their budget, originally estimated to support 5 hours of work, was completely depleted after only a few prompt interactions. The slow response from the OpenAI development team sparked intense protests from numerous developers on social media platforms. In response to user protests, Vaibhav, a member of the OpenAI technical team, first publicly collected user feedback identification codes. Subsequently, Sottiaux acknowledged that the anti-abuse system might have made a misjudgment and promised to follow up. In the latest quota reset announcement, Sottiaux emphasized that the team had implemented initial mitigation measures. Although internal investigations showed that the affected user base was not large, the official decision was still to reset the usage limits for all Codex users free of charge and to continue monitoring the system status.
Meta imposed restrictions on the use of Claude and Codex due to concerns about the risks of model distillation.
According to internal documents, Meta Platforms (META.O) has imposed restrictions on the use of Claude and Codex due to concerns about model distillation risks .(The Information)
Claude Code can also be compromised? Accused of using strangers' passwords to modify databases across servers.
According to Beating's monitoring, a user submitted a report on GitHub stating that when using Anthropic's command-line AI assistant Claude Code, server IP addresses, usernames, and root passwords that did not belong to the user appeared out of thin air in the AI's dialogue context. Subsequently, the local AI assistant directly read these passwords, automatically connected to someone else's server via SSH, and performed write modifications to the database. In other words, the user's AI, using someone else's account and password, mistakenly connected to and modified someone else's production database. Community technical staff analyzed that the root cause of the problem may be the failure of the "prompt prefix caching" isolation mechanism of the large model. To reduce computational costs and speed up the process, the cloud-based large model caches the preceding text of user dialogues; if the cache keys between different users collide or become confused, someone else's confidential cache may be incorrectly concatenated into your dialogue. If this conjecture is true, any developer using Claude Code faces the risk of their server usernames, passwords, and core source code being obfuscated and leaked to other users. However, it cannot be ruled out that this is merely a case of the model's illusion coincidentally guessing the real IP address and weak password, or that local project history contaminated the context. This issue has been tagged with the security category (area:security) by GitHub's automated system, and all parties are awaiting official verification and conclusions.