Dive Brief:
- Businesses looking to capitalize on AI should establish the technology as a new operating layer, restructure workflows, tie AI to business KPIs and redesign job roles, according to a July report from consulting firm West Monroe. The report draws on proprietary research from more than 400 U.S. business leaders.
- As executives reshape operations, they’re facing a lack of trust in the technology, rapid change as model releases outpace planning cycles, evolving regulatory risk and capacity constraints, the report said.
- Rather than bet on one platform, IT leaders should prepare data and infrastructure to be as flexible as possible, according to Erik Brown, senior partner of technology and experience at West Monroe and a report author. “Lock-in is super important to consider when it comes to the large AI vendors,” Brown told CIO Dive.
Dive Insight:
The AI market is evolving quickly, propelled by providers such as Anthropic and OpenAI rolling out new models and updates to existing releases. To avoid being left behind, CIOs should adapt planning cycles and technology, positioning environments for the modern pace of change.
By continuously assessing model releases, executives could discover AI tools that drastically reduce costs while providing similar outputs, Brown said. Surprise AI costs are already forcing enterprises to rethink how they implement AI, according to a cost governance report from Mavvrik’s.
CIOs should also look for vendors that offer flexibility and openness, Brown said. Particularly as system lock-in concerns rise, with 7 in 10 senior executives noting that it would be challenging to switch from their primary AI provider, according to IBM’s Institute for Business Value.
“We need that flexibility, we need the ability to experiment consistently across the board,” Brown said.
For IT leaders, the AI conversation has shifted from experimenting with the technology and figuring out where it will disrupt the business to looking for tangible returns on investment and demonstrating value to executives.
“To me that is the No. 1 theme we’re working with CIOs, CTOs, heads of product on,” Brown said. “How do we really measure the efficacy and make sure we’re using AI effectively.”
In some cases, effective AI use could mean assigning basic tasks to lower-level AI models while routing complex tasks to more powerful models. Measuring output is also critical, such as whether an enterprise is able to create a market-ready product with reduced, more efficient teams, and whether the actual product is more reliable, Brown said.
But even when the tools are in place and the metrics set, enterprises will still face challenges around employee buy-in, Brown added. Enterprises can often start with small use cases and teams, finding trusted people who are open to new ways of working and can act as influencers, he said.
“We are not getting replaced by AI,” Brown said. “Like most disruptive technologies, the way we work is going to drastically change. Some jobs will get replaced by AI, but that will allow humans to work more effectively. That being said, there’s human nature that’s part of this, and fear of, ‘What does this mean for me if I start automating my job away?’”
Lack of reliable data can also pose a challenge to enterprises as AI use cases get more complex, he added. Building data foundations alongside AI capabilities will help enterprises hit goals without slowing down innovation, Brown said.
