- Meta said that the time it took for its AI to run was linked to access to the internal system.
- The company said it was investigating an incident that it described as a configuration issue that led to the test being carried out. Meta added that the test was irregular.
- Meta said it is investigating potential issues with OpenAI and Anthropic in the United States, saying it is investigating a possible cause of the test being carried out.
Meta said that the time it took for the test to run was affected by internal system access. The company said that it is investigating how the issue happened and that it was caused by a configuration problem.
Meta said that its investigation found that the cause was misconfiguration, which, according to the company, led to the test being carried out. Meta also said that the test was irregular, which the company said it was investigating, and for comment BBC referred to Irregular.
Independent reports that Meta received internal access to the time it took for the test to run. BBC also said that the other two companies OpenAI and Anthropic also reported misconfiguration, which led to the test being carried out. Meta said that it is investigating whether the misconfiguration was the cause of the test being carried out.
Independent also reported that the letter to the US had been sent without AI. However, the group of researchers said that OpenAI was being investigated for the potential cause of the test being carried out. The report says that the letter was written in English by researchers at Hugging Face, and that Bill Gates said that the test was being carried out. The report says that Meta, Anthropic, OpenAI and Google are investigating whether the test was carried out.