Lately, we've been talking a lot about artificial intelligence (AI) and how it can support and elevate manufacturing functions.

The applications for AI are immense. Not only can it predict demand and potential breakdowns, but it can improve product development and product inspection. And pretty soon, with agentic AI, programs and machines will be capable of acting on the insights and recommendations they deliver.

Of course, there are endless AI solutions on the market. For me, this begs the question: What does AI adoption look like? Additionally, what are the barriers to adoption?

According to Veev, cost and security risks play a role, but data quality and integration are major concerns. In its recent "The State of AI in Consumer Goods" report, the cloud software provider, with help from a third party, surveyed 150 CPG senior quality and IT leaders in North America and Europe about their companies' AI adoption and their data structures.

The majority of respondents (nine in 10) say their companies are actively using AI or are conducting trials, pilots or evaluations. Additionally, three-quarters say they use AI for at least one business function. Predictive analytics carries the most interest, with just over half (52%) saying predictive analytics are the top priority for AI.

However, AI tools are useless without quality data. Survey respondents cited data silos as a significant roadblock, but through the three survey waves Veev conducted over six months, the percentage of leaders citing data silos and poor data quality as top hurdles dropped by 22 percentage points.

"Although they've made progress by consolidating systems and strengthening data practices, data silos still remain a top concern for consumer goods leaders," the report reads.

Integration is also a concern, since some manufacturing environments are "patchy, partial or fragile." However, nearly two-thirds (62%) of respondents say they are consolidating legacy platforms into a unified solution to capitalize on the benefits of AI.

"An integrated digital ecosystem is requisite for data to connect and flow across the business," the report reads. "When data and processes connect seamlessly across systems, leaders can derive quality insights quickly, improving decision-making."

Improved data quality and integration allow manufacturers to work toward implementing agentic AI — and many are already investigating it. In fact, one in five CPG leaders say they are "seriously exploring" agentic AI's potential and where it can deliver value.

Berenice Vettore, global chief quality officer for The Estee Lauder Companies, captures the possibilities the best: "We are maybe the last generation to have organizational charts with only humans," she says in the report. "Now we will have org charts with AI agents, and agents managing agents. It's the moment to dream. It's the moment to really say, ‘If everything is possible, what could we do?'"