Horváth & SANNOS AI partner on finance control review

Horváth and SANNOS AI have formed a partnership focused on using AI to assess financial control evidence, starting with financial controls.

The arrangement combines SANNOS AI's document assessment and compliance automation with Horváth's consulting and transformation expertise. Its aim is to link the review of policies, reports and control evidence with follow-up work to address gaps, improve controls and implement changes.

Finance teams and advisers often spend significant time reviewing documentation before they can examine exceptions or decide what action is needed. The partnership is intended to automate part of that assessment process while leaving final judgement to specialists.

SANNOS AI assesses documents against defined requirements and produces findings, gaps and proposed remediation linked to source pages. This approach is meant to help reviewers trace each finding back to the underlying material before reaching a conclusion.

Horváth helps organisations turn those findings into prioritised improvements, changes to operating models and practical implementation steps. The work begins in financial controls, but the scope also covers governance, risk and compliance, third-party risk, cybersecurity and operational resilience.

From review to action

The partnership addresses a common problem in finance transformation projects: identifying an issue is only part of the process. A control gap or compliance finding must still be understood in context, discussed with control owners and translated into a response that fits the business.

The value of automated document review therefore depends on what happens after the assessment stage. For CFOs and finance leaders considering AI projects, the central question is not just whether review work can be done faster, but whether the output can be checked, understood and used in decision-making.

The model could reduce the amount of specialist time spent on detailed document assessment. In turn, clients could devote more time to investigating exceptions, resolving weaknesses and strengthening controls.

Even so, it does not remove human oversight from the process. Expert validation remains part of the approach, and reviewers are still expected to decide whether a finding is relevant and whether any proposed response is appropriate.

Human judgement

The emphasis on human review reflects a wider issue for organisations adopting AI in regulated and control-heavy functions. Automation may reduce manual work, but finance leaders still need to know whether requirements are clearly defined, whether the evidence is suitable and whether each conclusion can be traced back to a source.

Other practical questions remain. Organisations need to decide how to handle missing information, how to address disagreements with the output and what additional context may be needed before a decision is made.

In that sense, the partnership is positioned less as a replacement for finance professionals than as a way to shift their time away from repetitive document review and towards judgement-based work. This includes discussions with control owners, evaluation of exceptions and implementation of remedial action.

Wider scope

Although financial controls are the starting point, the broader areas identified by the firms suggest they see similar demand across other parts of risk and compliance work. Governance, third-party risk, cybersecurity and operational resilience all rely on requirements, evidence and specialist interpretation, even if that evidence varies by use case.

This wider scope matters because many finance functions are under pressure to improve control environments while managing transformation programmes. Tools that can shorten evidence review without removing accountability may appeal to teams trying to balance compliance, efficiency and change.

For Horváth, the arrangement extends its advisory role into a workflow in which assessment outputs feed directly into control improvement and implementation work. For SANNOS AI, it places its technology within a consulting-led process that links findings and proposed remediation to business decisions.

The partnership centres on a practical handover point in finance transformation: the moment when an assessment finding must become an operational change. Rather than treating document analysis as an end in itself, the two firms are seeking to connect it to decisions on responsibilities, priorities and follow-through.

Experts validate the findings and retain final judgement.