# AI Won't Fix Enterprise Complexity, TIAA and Accenture Executives Say, Rewiring Will > TIAA's Sastry Durvasula and Accenture's Manish Sharma argue most companies are layering AI onto broken systems instead of fixing the underlying data and workflows first, according to Fortune. - Source: Money Standard - Canonical URL: https://moneystandard.co.uk/article/ai-wont-fix-enterprise-complexity-rewiring-will-tiaa-accenture - Author: Money Standard Newsroom - Section: Business - Published: 2026-08-22T01:47:22.291Z - Updated: 2026-08-22T01:47:22.291Z - Tags: Artificial Intelligence --- Most companies investing heavily in AI are not seeing much to show for it, and the problem is not the technology itself, according to TIAA chief operating officer Sastry Durvasula and Accenture chief strategy officer Manish Sharma, writing in Fortune. Their argument centers on a railroad metaphor: a company can buy the fastest trains in the world, but if it keeps running them on the same aging rails, the trains never actually go any faster. Only a fraction of businesses are turning AI investment into measurable results, the two executives write, because most organizations are stacking AI agents on top of decades of legacy IT systems, siloed data, and workflows that were already broken before AI arrived. Rather than speeding the business up, that approach just automates the existing dysfunction faster than before. The numbers back up how widespread the problem is. Only 5 percent of businesses report that their data is actually ready for AI, and Gartner projects that 60 percent of AI projects will be abandoned through 2026 largely because the underlying data was never in shape to support them. TIAA points to its own experience as a counterexample: after modernizing its systems first, the company saw a 13 percent increase in digital engagement, and an internal AI platform reached 85 percent daily adoption among employees. Durvasula and Sharma lay out five steps they say separate companies that succeed with AI from those that do not: modernize the digital core before layering AI on top of it, treat data readiness as a prerequisite rather than an afterthought by building unified and governed data platforms first, redesign workflows entirely instead of automating broken ones as they stand, keep humans in the loop for high trust decisions such as retirement savings advice, and build for resilience and governance with technology that stays flexible rather than locking a company into one vendor. Their broader point is that the companies winning with AI right now are not necessarily the ones with the biggest budgets or the fastest rollout, but the ones willing to do the less glamorous work first, cleaning up data, modernizing core systems, and rebuilding workflows around how AI actually functions, before expecting the technology itself to deliver results. --- Originally published by Money Standard. Free to cite with attribution and a link to https://moneystandard.co.uk/article/ai-wont-fix-enterprise-complexity-rewiring-will-tiaa-accenture.