Experts at OpenAI and Anthropic are calling out Elon Musk and xAI for refusing to publish any safety research \u2014 and for potentially not having done any at all.<\/p>\n
In the wake of Grok, xAI’s chatbot, calling itself “MechaHitler” and publicly spewing a ton of racist and antisemitic vitriol, safety experts at those rival firms are, as flagged by\u00a0TechCrunch<\/a>, extremely alarmed.<\/p>\n “I didn\u2019t want to post on Grok safety since I work at a competitor, but it\u2019s not about competition,” tweeted OpenAI safety researcher Boaz Barak<\/a>.<\/p>\n Barak, who also works as a Harvard computer scientist but is currently on leave to work at OpenAI, said that the scandal was “completely irresponsible” \u2014 and suggested that the “MechaHitler” incident may just be the tip of the iceberg.<\/p>\n “There is no system card, no information about any safety or dangerous capability [evaluations],” he said of Grok, noting that the chatbot “offers advice chemical weapons, drugs, or suicide methods” and suggesting that it’s “unclear if any safety training was done.”<\/p>\n “The ‘companion mode’ takes the worst issues we currently have for emotional dependencies and tries to amplify them,” Barak added.<\/p>\n Shouting out Anthropic, Google DeepMind, Meta, and China’s DeepSeek by name, the researcher noted that it’s customary for such AI labs to publish what’s referred to in the industry as system or model cards, which show how the people who made it evaluated it for safety. Musk’s xAI, meanwhile, has published no such information about Grok 4, its latest update to the chatbot.<\/p>\n “Even DeepSeek R1, which can be easily jailbroken, at least sometimes requires jailbreak,” Barak quipped.<\/p>\n The OpenAI researcher’s sentiments were echoed by Samuel Marks, who works in a similar capacity at Anthropic.<\/p>\n “xAI launched Grok 4 without any documentation of their safety testing,” Marks tweeted<\/a>. “This is reckless and breaks with industry best practices followed by other major AI labs.”<\/p>\n Acknowledging that Google and OpenAI both “have issues<\/a>” of their own when it comes to safety evaluations \u2014 a nod to recent fiascos involving those companies choosing<\/a> not to immediately release<\/a> their system cards \u2014 the researcher pointed out that in those cases, the safety testing at least had been undertaken.<\/p>\n “They at least do something, anything to assess safety pre-deployment and document findings,” Marks wrote. “xAI does not.”<\/p>\n Dan Hendrycks, an AI safety adviser at xAI, claimed on X<\/a> that it was “false” to suggest that Musk’s company didn’t do any “dangerous capability evals” \u2014 but as an anonymous researcher on the AI-focused LessWrong forum suggested, based on their own tests, that the chatbot appears to have “no meaningful safety guardrails<\/a>.”<\/p>\n It’s impossible to say whether or not xAI did any safety testing ahead of time \u2014 but as these researchers demonstrate, it doesn’t matter much without any model card release.<\/p>\n