7 in 10 enterprises expected to abandon vendor-built agentic AI by 2028

Gartner forecasts that by 2028, approximately 70% of businesses will discontinue vendor-built agentic AI systems. The consultancy refers to this approach as forward-deployed engineering (FDE), where vendor engineers collaborate with customers to create and implement software tailored to specific needs. While FDE may provide swift initial progress, it can result in customers becoming dependent on costly external expertise, as highlighted by Gartner.

The success of FDE hinges on establishing the right engagement structure, encompassing factors such as scope, incentives, governance, ownership, and exit strategies, according to Mukul Saha, a senior director analyst at Gartner. The most successful FDE engagements come with clear guidelines on governance, delivering business value, intellectual property ownership, joint project ownership, knowledge transfer, and a defined exit strategy from the outset, Saha added.

Many providers now employ the term "forward deployed" to describe various services, including implementation, professional services, solution engineering, or AI consulting. Some providers charge premium fees without delivering sufficient depth, program management, or change management, raising concerns about the value of these services.

Gartner cautions software engineering leaders to utilize FDE exclusively for projects demanding profound product expertise, rapid adaptation, or close integration between the vendor's technology and the customer's operational environment. Otherwise, businesses might experience rapid early progress without developing the necessary internal capabilities to take control, potentially leading to paying premium rates for work that traditional services or partner models could provide more cost-effectively and predictably.

Furthermore, Gartner projects that only less than 20% of FDE engagements will convert recurring customer requirements into features within the vendor's core product by 2028. The analyst warns of an increasing risk of "FDE washing," where conventional consulting services are marketed as specialized offerings. This is not the first time Gartner has warned about the potential costs and pitfalls of AI projects; earlier this year, it projected that at least half of generative AI projects will exceed their budgets due to poor architectural choices and insufficient operational know-how, while most organizations attempting to build custom models will ultimately abandon them for similar reasons.