The Hidden Cost of AI 'Forward Deployed Engineers'
AI-generated image · US National Wire
Vendors are embedding high-paid engineers into client operations to accelerate AI deployment, but Gartner warns that a lack of governance could lead to intellectual property leaks and vendor lock-in.
Enterprise software vendors are increasingly deploying "forward-deployed engineers" (FDEs) to help clients implement AI systems, a move that, as The Register first reported, accelerates production but introduces significant operational risks for the buyer.
FDEs act as consultants-without-portfolio who are embedded at client sites without being tied to a specific project. Their primary goal is to build valuable applications at speed using the vendor's technology. This model, which Palantir claimed to originate in 2007, has seen a recent surge in investment. AWS recently announced a $1 billion investment in a new FDE unit, while AI firm Anthropic has committed $100 million to training FDEs. Other major players heavily invested in the model include SAP, Salesforce, and Workday, the latter in partnership with Accenture.
For operators, the appeal is speed. Salesforce CEO Marc Benioff stated during a December earnings call that customers utilizing FDEs deployed the company's AI technology 33 percent faster than those who did not. This acceleration is critical as the tech industry seeks to recoup investments from a trillion-dollar AI spend reported by Gartner, which is projected to grow by 34.7 percent in 2026.
However, the lack of strict governance over these embedded resources can create strategic vulnerabilities. Mukul Saha, software engineering senior director at Gartner, told The Register that FDEs learn during implementation and take that knowledge back to the vendor to improve the product. Saha warned that there is a "fine line" here, noting that FDEs could potentially pass a user's competitive advantage to other clients, which would constitute a violation of intellectual property.
Beyond IP risks, Gartner warned that reliance on FDEs for every iteration can lead to massive vendor lock-in. If FDEs become the only individuals who understand the specific workflows they have built, the client is left in a precarious position. Saha advised clients to define their use cases before FDEs arrive and to establish a clear exit strategy for when those resources leave.
Financial risks also loom. Craig Guarente, founder and CEO of Palisade Compliance, told The Register that FDEs typically do not monitor licensing, tokenization, or commercial arrangements. Consequently, buyers with capped usage may unexpectedly hit limits and face additional costs, while those with open-ended contracts may be "really exposed" if they do not closely monitor the FDE's activities. Guarente added that FDEs may also share the client's IT strategy and software usage back with the home vendor.
Despite these risks, Gartner notes that FDE engagements can be valuable if they are governed by clear guidelines regarding IP ownership, knowledge transfer, project co-ownership, and business value delivery.