The Industrial Internet of Things (IIoT) requires sophisticated Intrusion Detection Systems (IDS) due to its numerous security vulnerabilities. By using Bidirectional Gated Recurrent Units (BiGRU) with an integrated attention mechanism, this research aims to improve IDS performance in IIoT contexts. The Bot-IoT dataset is used by the BiGRU-based intrusion detection system (IDS) to analyze a wide range of network traffic that represents both typical and attack scenarios, including internal threats. While the BiGRU model analyzes data both forward and backward to capture intricate temporal patterns and relationships, the attention mechanism highlights important aspects. Proposed model show that by increasing detection accuracy and decreasing false positives, the BiGRU with attention mechanism greatly enhances IIoT security, surpassing that of conventional methods.