Compliance teams using EQS Group's platform now have an AI layer, Q by EQS, operating directly inside the system of record with access to case data, disclosures, policy libraries, due diligence results and workflow history. AI-powered whistleblowing triage, case classification and policy assistance are available now, with agentic capabilities to follow later this year.
The launch draws on the EQS AI Benchmark Report Volume 2, which the company describes as the industry's first evaluation of AI performance on compliance work. It found frontier AI models achieve up to 87 per cent accuracy on real-world compliance tasks, with the top four models within a single percentage point of each other - making the surrounding expertise, workflows and governance the real differentiator.
Initial use cases include identifying involved parties in whistleblowing reports and automatically blocking their case access, report triage and severity assessment, investigation support including evidence summarisation, disclosure risk scoring, and natural language data queries with opt-in benchmarking against anonymised EQS peer averages.
"Building AI that works in compliance is not a model problem - it's a domain problem," said Moritz Homann, Head of AI at EQS. "That includes human oversight at important decision points and full auditability."
The planned agentic layer will execute multi-step tasks in the EQS Compliance Cockpit under human oversight. Compliance teams can configure where Q assists, where it is restricted and where it is disabled, and customer data is never used to train AI models operating in EQS products.
eqs.com