While many business functions have been quick to adopt AI, finance and accounting have been slower to do so.
Often, this reluctance is due to the fact that generic AI tools may not support compliance requirements. Generic AI models that promise the world are a non-starter in accounting.
Finance and accounting professionals need to be able to trust AI’s outputs, which means they need transparent AI models that allow them to share full audit trails with auditors and regulators. To move forward, CFOs need finance-specific AI that provides an unbreakable chain of thought and a clear audit trail.
Despite these reservations, the office of the CFO is increasingly under pressure to start using AI or at least exploring how it could support the department.
“Everyone’s being asked to do this, so nobody can sit on the sidelines with AI,” says Jeremy Ung, chief technology officer at BlackLine. The challenge organisations face is that they can’t treat AI in finance and accounting as a one-size-fits-all solution, as they might in other parts of the business.
When it comes to finance and accounting, the tolerance for inaccuracy is zero. It’s critical that AI is applied to these domains in a manner that’s trustworthy
“If you think about content creation, video animation or image generation, it’s ok for AI to make mistakes – it’s ok to have a six-fingered person or a dog with an ear missing,” says Ung. “For content creation, there’s a low cost to getting it wrong. When it comes to finance and accounting, the tolerance for inaccuracy is zero. It’s critical that AI is applied to these domains in a manner that’s trustworthy.”
This means building trust in AI outputs so finance and accounting teams feel comfortable that they are not being fed inaccurate information or having to spend an unnecessary amount of time correcting errors.
“It’s like swimming lessons,” says Ung. “You start in the shallow end of the pool with easy things, because you build your confidence that way. These days, we don’t throw people in the deep end right away. You start with areas where there’s more room to build that trust.”
For example, finance and accounting teams can build trust by using domain-specific AI technology to generate document summaries. This not only saves staff manual effort, but it also gives them an opportunity to review AI outputs and assess its accuracy in a lower-risk use case.
“To get to the AI future that we are seeing OpenAI, Anthropic and others present as a potential outcome, we need to slowly scaffold and build these steps to get finance professionals comfortable with how AI is going to work,” says Ung.
Finance and accounting teams can also grow confidence in AI by using tools that flag risks and highlight where humans may need to step in and review or focus their efforts. Teams can also use AI tools that rank AI’s own outputs with a confidence score, while also enabling finance professionals to adjust their risk tolerance around unsupervised AI use. At the same time, finance leaders need confidence that those outputs are auditable to maintain regulatory compliance.
“This is all about a chain of thought. It’s about showing the reasoning or the way in which an AI system or agent arrives at a particular conclusion and then documenting that,” says Ung.
Another key component of building trust is upskilling the existing workforce so they can work with AI in a collaborative manner, where it augments their jobs by handling more mundane tasks rather than replacing them entirely. This process is a challenge though, given that the technology is so new.
“This is not something you can just Google and find a ‘how to’ guide on, or watch a YouTube video. We are building playbooks and refining them, but we have to work with our customers on this,” says Ung. “It is a collaborative journey, because this is such a new domain. And it’s not necessarily about just building better AI systems. It’s about building a workforce that knows how to leverage AI tools.”
BlackLine’s vision for the future is a world where companies have a combined AI and human workforce that operates in a complementary way. This approach streamlines and optimises work, while allowing human workers to apply their domain expertise to more strategic matters that drive the business forward.
“Our CEO puts it this way: we’re probably the last generation to manage only humans and we’re going to be the first generation to manage a human plus agentic AI workforce,” says Ung. “So it’s important that you build trust and capability in the workforce to do this.”
BlackLine is helping organisations achieve this through Verity – a suite of AI capabilities that sit on the BlackLine platform and help finance leaders combine human and AI agents to enhance the effectiveness of the office of the CFO.
“We start with best-in-class AI models, then purpose-build and train them specifically for the rigours of finance and accounting,” says Ung. “So we’re giving them finance and accounting tools, but we’re giving them tools where you have an audit trail and a record of their activity that auditors already trust.”
Ung says there are four key areas where Verity can help finance departments transform the office of the CFO into a strategic engine for the wider organisation. First, Verity can generate intelligent insights for users by interrogating data in BlackLine’s platform and surfacing information that finance and accounting teams might not otherwise have had, such as flagging anomalies or risky transactions.
Second, it can support finance and accounting leaders with generative AI-related content, such as document summarisation and producing financial statement summaries.
“That’s where we can simplify things like variance analysis and create a default or preliminary explanation. Then users get the ability to review and refine that,” says Ung. “They can give it a thumbs up or a thumbs down and we then take those signals to refine our models further.”
With a digital workforce, we truly mean a future where AI agents augment teams. We’re using AI to simplify work that would otherwise fall to humans
Third, Verity enables greater process automation to help CFOs with digital finance transformation. For example, Verity allows users to manage growing data volumes by automatically handling tasks such as identifying unmatched transactions or handling remittance processes and identifying payment terms in invoices.
Finally, Verity can deliver agentic AI experiences that are designed to ease the workload on finance and accounting teams and give human workers the time to focus on more strategic matters.
“At BlackLine, we’re really creating a digital workforce – AI agents that can augment work and operate like people in your org chart,” says Ung.
For example, AI agents can perform preparation tasks that would otherwise take humans many hours to complete. The technology can also help support collections workflows, for example by automatically calling a customer to chase unpaid invoices. AI agents can then identify a propensity to pay and open a dispute workflow if needed, rather than a human having to place a call. In addition, Verity also has an AI-powered supervisory layer, Vera, which acts as a leader and coordinator for a team of agents, delegating complex tasks and directing workflows.
“These are not just point solutions; they are end-to-end workflows,” says Ung. “With a digital workforce, we truly mean a future where AI agents augment teams. We’re using AI to simplify work that would otherwise fall to humans, delivering time savings through these agentic experiences.
By embedding accounting expertise and auditability into the way Verity works, finance and accounting leaders can start to embrace AI tools with confidence and unlock future-ready financial options.
For more information, visit blackline.com
While many business functions have been quick to adopt AI, finance and accounting have been slower to do so.
Often, this reluctance is due to the fact that generic AI tools may not support compliance requirements. Generic AI models that promise the world are a non-starter in accounting.
Finance and accounting professionals need to be able to trust AI’s outputs, which means they need transparent AI models that allow them to share full audit trails with auditors and regulators. To move forward, CFOs need finance-specific AI that provides an unbreakable chain of thought and a clear audit trail.