The continued digitalisation of banking systems has raised a demand for more reliable data-based decision-making, in particular when referring to financial forecasts as covered by e-banking applications. This research also investigates the usage of AI-based decision-making systems to facilitate forecasting effectiveness in Jordanian commercial banks. Field research was carried out and 390 employees, working at 14 commercial banks in Jordan, responded to an organised questionnaire. Although the minimum required sample size was 384 respondents, a total of 390 valid responses were collected and used in the final analysis, thereby exceeding the minimum sample requirement. This research concentrates on three dominant categories of AI applications, including expert systems (ES), machine learning (ML), and Robotic Process Automation (RPA), which together are analysed for their effect on forecasting results in the context of customer churn, debt repayment, as well as investment analysis. The results of the multiple regression analysis indicate that AI applications contribute to improvements in forecasting accuracy, with machine learning and RPA showing relatively stronger effects. Expert systems were found to support investment analysis and debt repayment forecasting; however, their influence on customer churn prediction was more limited. In general, the findings indicate that AI applications are not confined to routine automation but are increasingly used as decision-support tools that assist financial analysis and forecasting activities in banking systems.

In this paper, a novel hybrid edge-detection steganography technique is proposed, which greatly increases the payload capacity without losing much of its invisibility. Conventional least significant bit (LSB) steganography has a low payload capacity and is sensitive to statistical analysis. Our approach combines Canny and Sobel edge-detection methods to find the optimal embedding regions and then performs Most Significant Bit (MSB) modifications in edge areas where the human visual system (HVS) is less sensitive to changes. Experimental results show that the performance of our proposed method outperforms conventional LSB-based steganographic methods by an average of 42.3% in payload capacity, while maintaining a PSNR greater than 38 dB and an SSIM above 0.95. The proposed method is also more robust against statistical attacks, such as chi-square analysis and RS steganalysis, which are critical challenges in secure data transmission.

This research examines file protection in cloud environments using the SHA-256 algorithm. It evaluates secure hash variants in file transfer systems, aiming to improve throughput while meeting security requirements. Data analysis involved gathering and cleaning datasets, measuring encryption times, and applying statistical and graphical methods. Results show SHA-256 as an effective base for encryption, with parallel processing increasing efficiency when real-time speed is essential. The study highlights the value of transformational techniques to boost performance and recommends hybrid systems that combine SHA-256 with other algorithms for stronger security. These outcomes support further work on cryptographic methods to strengthen the safety and reliability of cloud infrastructures.