The adoption of AI into the corporate world has surged this year, beyond initial expectations for 2024, with two thirds of workers now using their own AI tools for work.
This was a key finding in Forrester’s Q2 AI Pulse Survey, 2024, which shows that US companies still have many reservations about the use of generative AI (genAI), like hallucinations where AI delivers incorrect outputs, accessing good training, and complying with data governance and protection requirements.
“But BYOAI is not a strategy,” says Forrester. “Enterprise AI leaders now realise that for longer-term AI success, they need an effective strategy combining data and AI to ensure AI enters through the front door. This underlines the requirement for a strategy that can adapt to the swift pace of innovation.
“The strategy must include estimating business impact and ROI, selecting the right use cases, plans for cleaning and governing data, aligning on an operating model, training talent, experimenting with new application architectures, partnering internally and externally, and balancing risk and reward. “
“This will not happen overnight. In the year ahead, you’ll need to put your nose to the grindstone to develop an effective AI strategy and implementation plan. To succeed in 2025, companies will need strong leadership, fine-tuned strategies, and the reset of data and AI plans in line with their AI ambitions.
Forrester has made the following predictions for 2025:
– Most companies fixated on AI returns will cut back prematurely
AI is enabling companies to achieve better customer service, improved employee efficiency, and find new commercial opportunities. However, a rebalancing is on the way. Promising use cases that companies explored in 2023 are now the norm and part of business software. Business heads are also becoming aware that returns from AI investments will not arrive as quickly as they expected. As a result, there is a shift towards achieving those returns more gradually. In Forrester’s Q2 AI Pulse Survey, 2024, close to half of US genAI decision-makers said their companies anticipate financial returns on their AI in the next one to three years, while most of the rest believe it will happen in the next three to five years.
– Almost a third of enterprise CIOs will integrate CDOs
CEOs will rely on CIOs to spearhead AI technology initiatives to bridge the gap between business and technical knowledge. This is where Chief Data Officers (CDOs) come in. CDOs and other senior data and analytics leaders, are shapeshifters, equally at ease in IT and business departments. Forrester’s State Of Data, Analytics, Measurement, And Insights Survey, 2024, shows that over a third of respondents say that equal shares of the most senior data and analytics heads report to the CEO and CEO.
– Over a third of highly-regulated companies will merge data and AI governance
From February 2025, the EU AI act will fine companies up to 70% of their worldwide earnings for breaches and in the US, possible AI and tax penalties will also create the need for integrated governance. As AI is a data app and AI governance is essentially data governance, over a third of highly regulated enterprises will merge their data and governance programmes. To comply, companies need to understand how AI models use data and the type of data involved. But in the absence of clear guidelines, standards, or certifications, businesses must urgently partner with third parties to develop a new shared responsibility model and learn from past mistakes. AI governance is especially complex because of its rapid evolution and sophistication.
– Predictive AI will become the foremost focus for more than half of use cases
According to Forrester’s Q2:2024 AI Pulse Survey, 2024, almost equal numbers of use cases leverage genAI and predictive AI. But, this balance will shift in favour of predictive AI. As companies come up against challenges in applying genAI in ways that meet expectations, they will intensify their focus on predictive AI applications. Tried-and-tested predictive AI use cases like predictive maintenance, personalised customer experiences, supply chain optimisation, and demand forecasting will syphon more investment away from genAI in 2025.
– Three in four companies that go it alone on agentic architectures will fail
AI agentic architectures, a subfield of AI focused on creating autonomous systems, were a promising technology in 2024. But they’re not yet fully developed and it could take two more years to reach their ambitious automation dreams. With the aim of agentic AI to extend genAI beyond simple tasks, companies are excited about the prospects. But the reality is that the architectures are complex, requiring multiple models, advanced data architectures, and specialised know-how. Coordinating these models for specific outcomes is a challenge that will likely disappoint developers. As a result, most companies that try to build these agents themselves in 2025 will fail and will turn to consultancies to build custom agent setups or use agents setups offered in their existing software vendor ecosystems.