What does it take for accountants to trust AI with work that has to be right?

In Argentina, Brazil and Chile, the answer starts with the market itself. Tax rules change quickly. Financial workflows are fragmented. Compliance leaves little room for error.

That is why AI in accounting software cannot be generic. For Xubio, Conta Azul and Laudus, useful AI is being built close to the realities accountants and businesses face every day. Across Visma, those local lessons can be strengthened by shared product thinking, security standards and technical expertise, without losing the context that makes them valuable.‍

Complexity sets the brief

Latin America does not offer easy conditions for accounting software. Regulatory change is frequent, tax systems are demanding, deadlines are unforgiving, and in some markets, inflation adds another layer of pressure to everyday financial work.

That context matters because it changes what AI is for. In these environments, artificial intelligence is not a decorative feature or a promise of autonomous finance. Its value lies in reducing manual work, improving reliability and helping accountants and business owners manage complexity with less friction.

Across Visma, this balance between local autonomy and shared capability is what makes the difference. Each company stays close to the customers, regulations and workflows of its own market, while product thinking, AI practices, security standards and technical learnings can travel across the group.

The result is not centralised AI for every market. It is AI shaped locally, strengthened collectively, and built for the complexity of mission-critical work.

Xubio: helping accountants cope with Argentina’s regulatory churn

In Argentina, the pressure comes first from volatility and regulation. Tax rules change often, reporting obligations are spread across national, provincial and municipal levels, and accountants spend a significant amount of time simply keeping systems aligned with the latest requirements.

Marianela Conde, Product Manager at Xubio, puts it plainly. “In Argentina, the biggest challenge is not adding new features, it’s keeping the system aligned with regulatory changes.” That pressure is especially acute when authorities alter reporting formats or invoicing requirements, and software providers have to respond immediately.

For accountants, that complexity creates a steady burden of manual work. Transactions need to be categorised, reconciliations checked, and supporting documentation prepared, with many firms still relying on Excel to organise information before filing. The risk of human error remains high, and mistakes can have regulatory consequences.

This is where AI has immediate practical value. Xubio is using it to support transaction categorisation, simplify reconciliations, and reduce repetitive processing. 

The next step is moving towards automatic interpretation of bank statements and more conversational ways of retrieving business information. In Argentina, AI is being applied to relieve the administrative strain that stops accountants from spending more time on advisory work.

Conde is clear that the professional role remains intact. AI can prepare and organise information, but judgement still sits with the accountant. In a market this complex, that is less a limitation than a realistic design choice.

Conta Azul: organising Brazil’s fragmented financial workflows

Brazil’s challenge is different. Its tax system is notoriously complex, shaped by layers of federal, state and municipal rules, and is now undergoing a major reform that is changing how companies manage compliance. At the same time, financial information often moves between small businesses, accountants, banks, tax documents and specialist systems, creating a need for better connection and structure. 

For Conta Azul, this gives AI a clear direction.  AI has to do more than read documents. It has to connect disconnected parts of the workflow.

Alex Corcioli, Chief Product Officer at Conta Azul, says the opportunity is not simply to layer intelligence on top of the product. It is to simplify the movement of financial information through channels that are messy, document-heavy and often poorly structured.

The company is also getting proactive — helping customers spot cash flow, inventory or tax issues before they become problems.

This reflects a broader product belief. In financial software, users are often less interested in conversation than in action. They want systems that help them move faster, reduce pressure and make better decisions without asking them to trust a black box.

In a market as demanding as Brazil, AI earns its value by bringing structure to complexity.

Laudus: building intelligence in layers in Chile’s ERP market

In Chile, Laudus is building AI with a clear sense of sequence. Its AI strategy has been built in layers, beginning with assisted intelligence rather than autonomous decision-making.

The first visible layer is a conversational assistant that lets users generate financial reports through natural language prompts. They can ask for averages, rankings, cost-centre breakdowns or summary views without exporting data or building formulas manually. “This is not about replacing the accountant,” says Jaime Sanz, CTO of Laudus. “It’s more like giving them a junior assistant who prepares the repetitive work so they only have to validate it.”

That fits a market where errors carry real tax and legal consequences, where trust means speed and usefulness, without losing professional oversight.

The company is now moving into financial analysis, evaluating liquidity, profitability and solvency to flag risk before users go looking for it.

Laudus also operates under frequent regulatory change, forcing urgent updates to declarations, invoice formats and certification processes. That shapes priorities, and explains AI investment in modernising legacy code and preparing its architecture for a more agent-based future.

What the region shows about useful AI in accounting 

Taken together, the three companies reveal a clear regional pattern. In Argentina, AI is easing the manual burden created by constant regulatory change. In Brazil, it is bringing structure to fragmented financial operations. In Chile, it is being introduced in carefully managed layers, improving access to insight while respecting the demands of compliance.

The broader story, however, extends beyond Latin America. It shows how useful AI can scale without becoming generic by remaining close to local complexity, sharing proven approaches across the group, and keeping trust at the centre of mission-critical software.

The common thread is discipline and proximity. Each company is starting from the realities of its own market and applying AI where accountants and businesses feel the pressure most. The result is something more credible and durable. They are building technology that supports professional judgement, reduces friction and turns local knowledge into collective progress. 

‍