David Moscatelli, CEO of Go Abacus, says local AI adoption is gaining traction in highly regulated industries such as banking and healthcare, where data privacy concerns are pushing organizations away from public AI providers. According to him, many institutions are becoming more reluctant to send sensitive information to external platforms and are instead prioritizing systems that can run inside their own environments.
This shift is being driven by two main factors: tighter control over private data and better visibility into operating costs. Moscatelli said organizations want AI tools that allow them to keep information on-premises while avoiding the uncertainty that can come with relying on third-party public services.
Privacy and control are becoming central requirements
In sectors handling sensitive financial and medical data, compliance and security requirements often shape technology decisions. Local AI is increasingly seen as a practical option because it allows institutions to process information within their own infrastructure rather than exposing it to outside providers. That makes privacy protection a core selling point for on-premises deployment.
Cost management is another reason local systems are attracting attention. For organizations with large-scale and recurring AI workloads, running models internally may offer a more predictable framework for long-term deployment and budgeting.
Go One targets on-premises AI at scale
Go Abacus said its Go One device is designed to support on-premises AI by connecting to employee PCs and can handle up to 2,000 concurrent users. The system can also be scaled further through daisy-chaining, giving organizations a way to expand capacity without shifting away from local deployment.
The company added that Go One OS uses specialized language models trained on client-specific data, aiming to improve efficiency and deliver more deterministic results for targeted tasks. Go Abacus also said stress testing has significantly reduced the risk of system failures, strengthening software reliability. Moscatelli highlighted the firm’s client-focused approach, including nightly updates to model weights, as part of its effort to balance privacy, adaptability, and performance in sensitive industries.