The recently released Predictions 2019 guide by research firm Forrester included an epitaph of sorts for 2018, “a year when reality chewed up the transformational ideal.” And what has more potential to be the transformational ideal than artificial intelligence?

Everyone’s talking about it, and many are dreaming big. Banks, FIs and fintechs alike are putting certain AI applications to use already.

But Forrester identified three key items that held the technology back in 2018:

• AI had insufficient information architecture, due largely to firms struggling with basic data governance issues
• AI was too horizontal, meaning its use was non-specific and minimally applied to firms’ operations; and
• AI was too confusing, limiting executives’ abilities to understand and trust AI’s return on investment in terms of operations or impact on customer experience.

In a financial services brief accompanying the guide, Forrester predicted AI and automation will drive “increased operational efficiency” in 2019.

“Executives will scale back their expectations of artificial intelligence transforming the entire business. Instead, firms will seek incremental gains by automating back-end processes, like anti-money-laundering, claims processing, compliance monitoring, liquidity management, and regulatory reporting, to eliminate paper processing costs and reduce manual errors. Firms will then look to streamline the routine work of frontline teams, creating more time for employees to create value for customers and apply AI to seek further efficiency improvements. Emerging AI-powered banking applications will help smaller banks to adopt these technologies.”

When it comes to explaining why financial firms are expected to take a turn toward caution, the brief puts the blame largely on “the looming threat of recession.”

“Rising borrowing levels and the trade shocks from populist politics make a volatile combination that threatens to tip major economies into recession,” the brief said. “That fear will make most banks, investment firms, and insurance companies cautious. Margin pressure in core businesses will spark budget cuts at many financial firms in 2019. Projects with no imminent return on investment will be first to go.”

Despite some likely challenges ahead, it’s pretty safe to say AI is more than a trend to be derailed by economic headwinds.

Deloitte’s recent State of the AI in the Enterprise report, a survey of more than 1,100 IT workers and executives from U.S.-based companies, found about 42% of executives believe AI will be of “critical importance” within 2 years.

About 82% of respondents said they’ve already seen a financial return from their AI investments, although the financial services and insurance industry has seen a less than median return (about 16%) on investment compared with other industries.

Considering the potential for growth, however, executives like Emily Steele, president for North America at software provider Temenos, have said AI can become a “backbone” for digital banking in the next five years.

A Bank Innovation survey conducted between August and September found 32% of 294 respondents who work at banks and fintechs ranked AI and machine learning as their top innovation priority.