The first thing to address is how technology is delivered. The history of IT "failure" is significant. Although there have been improvements, the core reasons are consistent. They primarily fall into four categories - basic planning, communication, decision making, and technology. The larger the project the more susceptible to the issues. This is why agile and continuous improvement are so attractive.
The slide is telling
This slide (although a few years old) showed that, even with the advent of certified project management, experts in architecture, and agile methods, project delivery has not gotten significantly better. It further shows that the larger the project the less likely for success.
Although there may be great improvements on the horizon (AI design, AI code, AI testing), faster delivery does not always translate into better products. In the real world where not everything is a green field and there is potential for impact on current operations. With current stakeholders (some with critical needs) "fail fast" does not work. It could cost lives.
The following word cloud was derived from analyzing why projects failed.
The same reasons have existed for years. It is planning, communication, decision making, and technology. We have all seen these excuses for project failure (and even used some).
These problems exist even in an agile world. It is most evident when agile is not based on a robust technical foundation. In order to meet sprints, teams are often encouraged to be creative (heroic programming). This leads to either refactoring in the future, or a solution that is hard to support.
Agile weaknesses get exposed in a solutioning environment. Where every problem is thought to be unique. Where every project seems to have specialized requirements. Although agile is designed to facilitate change, solutioning can make those changes localized and harder to support.
Agile thrives in an API/services/Product environments where common consumption patterns serve as the foundation (The Right Strategy). Every project, every sprint is delivery of capabilities within products (EA Diagram). Some capabilities may take longer to develop than a sprint, but they can often be developed as a capability within a product independently (new features). All changes follow a continuous improvement trajectory.
The product delivery philosophy empowers SDLC business partnerships.
By standardizing change (capabilities delivered within an EA domain product), delivery can be optimized and unleashed.