RAIDS AI unveils beta for real-time AI safety monitoring tool

RAIDS AI has announced the beta launch of its AI safety monitoring platform after completing a successful pilot phase.

The RAIDS AI platform has been developed to monitor artificial intelligence systems in real time, identifying and flagging rogue behaviour before such deviations can cause operational failures, introduce bias, or lead to regulatory breaches.

This new beta phase will enable a wider range of organisations to access and test the platform, allowing RAIDS AI to refine and further develop its offering before a full market release. Participating organisations during the beta period will receive complimentary access to the full range of RAIDS AI features for a limited time.

AI safety and compliance

The launch comes at a time when regulatory environments for artificial intelligence are becoming more stringent, particularly with the European Union's Artificial Intelligence Act. The Act, which established a comprehensive legal framework for AI, came into force in August 2024. Most of its provisions will apply from August 2026, impacting providers, deployers, and manufacturers of AI systems across various sectors.

RAIDS AI claims its platform constitutes essential infrastructure for compliance with the EU AI Act, helping organisations meet regulatory requirements through real-time monitoring and documentation of AI behaviour.

Findings and pilot phase

According to research conducted by the RAIDS AI team, more than 40 documented cases of historical AI failures have been found across different industries. Noted instances range from incorrectly generated legal citations, faults in autonomous vehicles, to the creation of unwarranted retail discount policies.

The company notes that the consequences of such failures often go beyond operational setbacks, with organisations incurring significant financial losses, regulatory fines, and reputational damage. The platform aims to address these issues by enabling early detection and management of problematic AI behaviours.

During the pilot phase, participants were provided with access to a real-time monitoring dashboard, behavioural alerts, incident logging functionality, and tailored safety reports. The RAIDS AI team supported pilot users throughout, with feedback from this phase set to inform further development in the beta.

Executive perspective

Nik Kairinos, Chief Executive Officer and Co-founder of RAIDS AI, commented, "This is start of the next exciting phase of the RAIDS AI journey. What the world can achieve with AI innovation is incredibly exciting, and no one knows exactly what the limits of it are. But this continued revolution must be balanced with regulation and safety."

"In all my decades of working in AI and deep learning, I've only recently become scared by what AI can do. That's because perpetual self-improvement changed the rules of the game. It's absolutely critical that organisations - their CIOs and CTOs - understand the severity of the risk. AI safety is attainable; failure is not random or unpredictable and, by understanding how AI fails, we can give organisations the tools to ensure they can capitalise on AI's ever-changing capabilities in a safe and managed way."

Product details and next steps

The RAIDS AI platform is co-founded by Nik Kairinos and Brett Stonefield and designed to alert users to real-time deviations in AI behaviour. Functionality includes live monitoring, automated alerting for potentially rogue actions, incident documentation, and the provision of reports tailored to specific systems and use cases.

Organisations are now invited to take part in the beta phase and trial the RAIDS AI platform at no cost during an initial period, with the aim of broadening the base of user feedback prior to wider rollout.