Up to 82% of AI Engineering Spend Lost to Bugs, Rewrites, and Delays: Study Finds

A wave of mid-2026 data shows artificial intelligence (AI) tools generate heavy hidden costs, with up to 82% of enterprise AI spending lost to bug fixes, rewrites, and review delays before reaching production.

The strain shows across three fronts, with engineering teams shipping unreliable code, Oracle taking on heavy debt to build AI capacity, and OKX restructuring how it evaluates talent in an AI-first workplace.

The Cost of AI-Generated Code

Entelligence AI surveyed 2,444 companies and found that for every $1 spent on AI tokens, $0.44 covers bug fixes, $0.27 rewrites AI-generated code, and $0.11 vanishes into review and merge delays.

Lightrun’s 2026 State of AI-Powered Engineering Report adds that 43% of AI-generated code still requires manual debugging in production after passing quality checks.

No engineering leader surveyed expressed full confidence in deployed output, a pattern echoed in Coinbase’s AI rollout and Cardano’s AI code split.

Oracle’s Leveraged Bet

Oracle has accumulated roughly $108 billion in total debt while raising another $50 billion in 2026 through debt and equity to fund AI data center buildouts.

🚨 ORACLE BET ITS ENTIRE FUTURE ON AI AND IS NOW DROWNING IN $108 BILLION OF DEBT.The chart tells exactly how this happened. In June 2025, Oracle jumped 13% on earnings. In September 2025, it exploded 41% in a single session to an all time high of $345 as Wall Street priced… pic.twitter.com/n8nBe8PJ5e

— Bull Theory (@BullTheoryio) May 28, 2026

Free cash flow sits near negative $13 billion. Over $300 billion of Oracle’s $553 billion backlog ties to OpenAI alone, a client that lost about $14 billion last year.

The exposure tracks with broader warnings about the enterprise AI cost crisis and the AI revenue bubble. Oracle’s June 16 earnings will test whether the bet on AI demand holds.

The Talent Reset

OKX CEO Stax Xu argued that AI agents accelerate execution while exposing workers who rely on impression management rather than outcomes.

The exchange now ties employee evaluations to AI proficiency, joining a wave of exchange AI mandates across the sector.

“It’s not AI that fundamentally changes layoffs. It’s that the AI era fundamentally changes talent requirements,” Stax Xu highlighted.

The data argues AI delivers real capability, but the operational, financial, and organizational costs are arriving faster than markets priced in.

Whether June earnings and engineering metrics narrow the gap will shape the rest of the cycle.