The Globe and Mail reports in its Wednesday, Aug. 26, edition that in mid-July, while testing new artificial intelligence technologies, Microsoft-backed OpenAI's powerful systems hacked into the on-line service Hugging Face after breaking out of their digital containers. A New York Times dispatch to The Globe reports that many researchers believe the incident showed that AI technologies are becoming dangerous and that leading labs must control their use. A month later, Z.ai, a Chinese lab, is set to release similar AI technology as "open weight" software, allowing anyone to use it freely. The upcoming release of China's GLM 5.3 on Friday highlights a longstanding debate among AI researchers. Some experts argue that the latest AI technologies are too risky for public access, while others believe open sharing is the safest approach. Dan Lahav, head of cybersecurity firm Irregular, emphasized the importance of open weight models in this discussion. The disclosure that OpenAI's technologies had unexpectedly hacked into Hugging Face, confirmed what cybersecurity experts have been saying for months: The leading AI systems are now shockingly good at identifying and exploiting vulnerabilities in computer software.