French AI newcomer Mistral AI recently launched a new content moderation API service, highlighting its ambition in the global AI field and demonstrating deep thought on AI safety governance. This moderation system, built on the Ministral8B model, can accurately identify nine categories of potentially harmful content, including sexual content, hate speech, violence, dangerous activities, and personal identity information. Notably, the API supports 11 major languages, including Arabic, Chinese, English, French, and more.

EY has successfully utilized artificial intelligence to detect audit fraud. The company tested a new AI system on the accounts of its UK audit clients, which identified two suspicious activities among the first ten companies tested. This demonstrates the potential of AI in enhancing audit quality and reducing workload; however, there are differing opinions among auditors on the level of reliance on AI, and regulatory bodies need to provide their views.

SALMONN is an audio-text multimodal large language model framework designed to expand the understanding and processing capabilities of large language models in the general auditory domain. The framework integrates components such as non-speech BEATs audio encoders, the OpenAI Whisper framework's speech encoders, and window-level Q-Former, achieving high levels of temporal resolution for audio-text alignment. After the activation adjustment phase, SALMONN has achieved competitive performance in tasks such as audio captioning and speech translation, demonstrating general auditory capabilities.