US companies have invested between $35 billion and $40 billion in generative AI (genAI) projects, but most efforts are stuck in the pilot stage, according to a report from MIT’s NANDA initiative. Only about 5% of the efforts lead to rapid revenue growth; the majority produce little or no impact, Fortune reports.

The core problem is apparently not the quality of the models being used, but a lack of integration, learning and alignment with corporate workflows. Companies often invest in sales and marketing solutions, but the biggest returns seem to be in back-office automation and streamlining internal processes.

The report also found that successful companies tend to buy specialized solutions and build partnerships, while in-house development projects fail significantly more often.