Harnessing AI for Cybersecurity: Real-Time Detection and Mitigation of Online Threats- A Survey

Now a days dues to huge impact of digitalization, all are using social media, internet which directly impact on day to day life increased reliance on digital communication. The rising threat landscape is characterized by an increased reliance on digital communication, making it a prime target for cyber-attacks, with 90% of attacks beginning with a phishing attempt, often through a malicious link sent to a victim By using shared infrastructure, Complex configurations, Third-party dependencies Artificial intelligence .this directly opens door for cyber-attacks. Cyber threats are becoming more sophisticated, targeting IT systems across industries Traditional rule-based cybersecurity measures often fail to detect dynamic threats effectively. These measures lack adaptability and rely on predefined patterns, limiting detection. Deep learning is effective in analyzing vast data and recognizing complex patterns Integration of advanced technologies like AI is essential to enhance cybersecurity.(AI) plays an important role in today's digital world, mainly in enlightening Cybersecurity and threat detection systems. This study presents an AI-powered intelligent threat detection method uses deep learning methods to recognize as well as neutralize a variety of online dangers. With an emphasis on their effects on security, dependability, and data integrity, it examines popular attack vectors such as malware, denial-of-service (DoS) assaults, phishing, SQL injections, and cross-site scripting (XSS). Conventional cyber security defenses often fail to keep up with these quickly changing threats. We perform a systematic review using the PRISMA method has provided new insights into cybersecurity. It helps better understand various technologies used to secure against threats and improve cybersecurity practices.