The State of Generative AI 2025: 3 questions to understand your agentic AI readiness

It’s been incredible measuring the adoption rate, wins, and concerns around AI and generative AI over the past three years. As soon as ChatGPT came onto the scene in late 2022, the team at Lucidworks knew we needed to capture these insights from people at businesses with active AI initiatives. That was at the same time that we began to adapt our AI-powered search solutions to the era of conversational, generative AI. It’s how Lucidworks AI was born. 

Now, we’re approaching the era of agentic AI readiness. Why use the term “readiness”? Well, our most recent research found that only 6% of commerce businesses have partially deployed an AI agent on their websites. But while most aren’t there yet, they are certainly preparing for it. Hence, getting ready.

So, in addition to continuing the important work for surveying people who actively work on AI projects to understand their hopes and fears, we went a little further this year. We deployed a first-of-its-kind agentic AI bot to autonomously assess the AI readiness of 1,100+ company websites. We had it look at 24 capabilities, from basic relevancy ranking to advanced AI-powered guided selling. The insights were fascinating. 

We’ve already had incredible momentum and excitement from releasing the findings of our report. Check out these Forbes and CNBC articles. Now that the 2025 State of Generative AI report has been out for a minute, we felt it was time to review the highlights. 

There’s a lot of goodness in this year’s report, and you can review it all right here.

What we asked our survey respondents

We’ve always been interested in capturing the technical aptitude of business users at Lucidworks. It helps us develop products that actually help people do their jobs better. It also educates the market on the reality of AI adoption, because we are committed to helping organizations make informed decisions about tech investments.

This year, we asked over 1,600 AI practitioners:

1. Do you plan on increasing your AI spend in the next 12 months?
2. Have your revenue, productivity, and customer/employee satisfaction benefited from AI?
3. Rate your level of AI concern across: cost of deployment, job displacement, decision transparency, data security, and response accuracy. 
4. How many models do you use, and which types?

And then, because we wanted to get in on the agentic AI game, we built our own AI agent called Guydbot, to compare those insights on what people told us to show us how effectively companies are using AI on their public-facing websites in their digital experiences. How did it work? Guydbot was assigned a task to autonomously analyze 1,100+ company websites across 24 capabilities. That’s where some of the really good stuff was found.

So, the main question we were curious about: “How many AI capabilities have commerce companies actually deployed in 2025?”

A few themes that emerged from the data

We got a bunch of interesting data, big and small, from the AI practitioners who took our survey. And then pairing that insights from our market intelligence AI agent Guydbot took everything to the next level.

1. “What percentage of companies are using agentic AI in 2025?”

What we found in the Lucidworks 2025 State of Generative AI in Global Business report was that only 6% of e-commerce companies have fully deployed agentic AI on their website. That was paired with more information that only 35% of companies have the foundations to be ready for effective agentic AI, meaning they had data and search essentials.

We also heard some alarming, but understandable, concerns from AI leaders in terms of deployment cost, response accuracy, and data security: 83% of leaders report anxiety around their AI projects. That’s an 8X increase in concerns since Lucidworks started the study in 2023.

Our take? The AI hype cycle makes business leaders feel perpetually behind in their AI endeavors. Our research found some of that to be true – that many companies lack agentic AI readiness because they don’t have the foundations, like hybrid search or language options. As Lucidworks’ CEO said: “Our report finds that too many e-commerce companies are trying to build Formula 1 racers around go-kart engines—and they might not even have enough gas to fill their tanks.”

But that doesn’t mean companies are (or should be) halting AI projects…

2. “Which industries are increasing AI spending in 2025?”

Overall, spending is stable compared to last year – around 62% of all AI leaders surveyed are increasing their investments in AI in 2025. That’s about the same as 2024, with 63% investing more in AI. But nothing compares to the massive interest at the beginning of the generative AI hype cycle in 2023, with 92% of all companies surveyed reporting increased AI investments.

When we narrow in on industries, the 2025 Lucidworks study found that only B2B companies are increasing their AI spend more than last year:

• B2B: 68% plan to increase AI spending in 2025, up from 63% in 2024
• B2C: 58% plan to increase AI spending in 2025, down from 61% in 2024
• Financial Services: 55% plan to increase AI spending in 2025, down from 70% in 2024
• Healthcare: 44% plan to increase AI spending in 2025, down from 51% in 2024

Our take? B2B has a lot of catching up to do in terms of modernizing their digital experiences, but they’re also implementing AI in a more “foundation-first” approach. B2B companies are less likely to have implemented as many AI capabilities as their B2C counterparts, according to Lucidworks’ Guydbot research. The research also found that 37% of B2B companies are realizing the most benefits from AI, only second to healthcare at 43%. B2C (33%) and FinServ (32%) are not seeing AI benefits as clearly.

3. “What AI capabilities do top retail companies have?”

We’re not fans of gatekeeping valuable information. Here is every capability – and partial or full adoption rate – for B2C retailers, according to Lucidworks’ 2025 AI Benchmark Report:

1. Relevancy Ranking – 74%
2. Dynamic Facets – 70%
3. Request to Sign-Up – 66%
4. Product Content Enrichment – 63%
5. Vector Hybrid Search – 61%
6. Null Response Avoidance – 56%
7. Typeahead – 53%
8. Custom Catalog – 52%
9. Product Availability Info – 51%
10. Customized Promotions – 50%
11. Personalization – 49%
12. Custom Recommendations – 48%
13. Search Results Content Integration – 47%
14. A/B Testing – 46%
15. Language Options – 38%
16. Customized Landing Page – 38%
17. Dynamic Pricing – 38%
18. Personalized Fitment – 37%
19. Guided Selling – 34%
20. AI Full-Service Chatbot – 19%
21. Interactive Product Q&A – 16%
22. Technical Interrogation – 13%
23. Conversational Commerce – 13%
24. Synthesis & Summarization – 11%

Curious what each of these capabilities looks like in action? Check out this blog with the 8 essential AI capabilities every business needs in 2025 (that most already have).

For data leaders, it’s a balancing act

This annual benchmark report is meant to be exactly that: a way for companies to benchmark themselves against others in their industry. It should validate what you’re seeing within your own organization, and possibly show you precisely where you should invest more to realize more benefits in AI. 

If you’re reimagining your digital experiences, the Lucidworks AI Benchmark Report has detailed insights on over 1,100 companies. You don’t want to miss it as you plan for 2026. It’ll be here before we know it. 

You can read the full report here.

Are you in B2B? Check out our new B2B infographic.

Are you in B2C. Check out our new B2C infographic.
We hope you get as much out of this data as we do — and as always, we’re here to help set the record straight on the state of generative AI in global business. It’s the dawn of the agentic AI era.