Wharton School researchers found managers blame the deploying employee, their manager or the risk team for a costly AI agent error at about a quarter each. Only about 10% pointed at the vendor, so roughly 75% of the blame stays inside the company with no agreed owner.
Perspective Coverage
3 publishers
- Builder
- Builder 17%
- Operator
- Operator 60%
- Investor
- Investor 23%
Reality
- Evidence50
- Adoption
- Not scored
- Hype gap+15
- Incentives55
- Confidence60
Wharton and GBK Collective found 28% of surveyed enterprise leaders report strongly positive AI returns, against 80% reporting positive ones. Departments tracking growth outcomes report strong returns at 35%, against 13% for efficiency-only ones.
Reality
- Evidence45
- Adoption70
- Hype gap+10
- Incentives50
- Confidence62
Wharton economists model AI traders that coordinate through prices without exchanging a message
Wharton finance professors Itay Goldstein and Winston Dou argue that agentic AI traders can coordinate through prices alone, with no messages exchanged. They say the result could be bubbles that hit retail investors especially hard.
Reality
- Evidence35
- Adoption
- Not scored
- Hype gap+25
- Incentives40
- Confidence50
Givers set the minimum for a small gift about 70% above what recipients would take
People in a Wharton-led study set the smallest worthwhile gift about 70% above what recipients would accept, so many holding only a small sum gave nothing. For anyone running a giving drive, the authors' field test shows a lower ask brings in more contributors while pulling gifts down toward the amount asked.
Reality
- Evidence55
- Adoption
- Not scored
- Hype gap+20
- Incentives45
- Confidence50
Chatbot users deferred to AI answers 80% of the time in a study aired at the New York Fed
Wharton researcher Steven Shaw told a New York Fed conference that chatbot users took the AI's answer 80% of the time and felt surer even when it was wrong. For banks, the fix discussed there is deliberate friction, paid for in the speed the tools were bought for.
Reality
- Evidence38
- Adoption
- Not scored
- Hype gap+20
- Incentives30
- Confidence40
Interviews with 760 finance employees tie remote work to slower help between colleagues
Peter Cappelli and Y. Jasmine Wu found remote work wearing down help between colleagues in interviews with about 760 employees of a financial firm. The cost lands on newcomers with no one to call, outside the individual deliverables supervisors were told to watch.
Reality
- Evidence40
- Adoption
- Not scored
- Hype gap+25
- Incentives35
- Confidence45
Cappelli's study of 760 employees traces a two-tier workforce to remote work
Wharton's Peter Cappelli and Jasmine Wu interviewed 760 staff at a global financial services firm and found remote and hybrid work cut colleague contact. The evidence comes from one firm, but it gives managers who treat hybrid work as settled a reason to re-examine what it costs in collaboration and training.
Reality
- Evidence38
- Adoption
- Not scored
- Hype gap+30
- Incentives55
- Confidence45
Wharton-CPA primer urges boards to take direct charge of corporate political spending
Public companies and their trade groups supplied over 40% of the $2.5 billion raised by party 527 committees in 25 years, a Wharton-CPA primer finds. Its authors want boards to govern that giving on purpose before possible post-midterm investigations.
Reality
- Evidence35
- Adoption
- Not scored
- Hype gap+20
- Incentives70
- Confidence40
Spending fell further than footfall in the neighborhoods ICE raided
Exequiel Hernandez matched 5,388 geocoded ICE raids to card transactions at 1.1 million locations and found weekly spending down 6.18% against a 2.73% drop in visits, with no recovery over time and no move to delivery.
Reality
- Evidence55
- Adoption30
- Hype gap+25
- Incentives60
- Confidence45
Labor economist says subtle heat effects on workers cost the economy trillions of dollars a year
R. Jisung Park says the damage from warming shows up as more on-the-job injuries, delayed trains and lost productivity on days too ordinary to make the news, and some of those injuries happen indoors.
Reality
- Evidence42
- Adoption
- Not scored
- Hype gap+15
- Incentives62
- Confidence52
The Bank of England sorted firms' AI into fraud defence and low-risk help for years
Sarah Breeden, the Bank of England's deputy governor for financial stability, has given a public account of a long and benign period of AI use inside London firms. She is now watching agentic systems.
Reality
- Evidence58
- Adoption32
- Hype gap+12
- Incentives55
- Confidence55
Wall Street read a yield target into Bessent's Treasury buyback move, economists say
Treasury's repurchase operations rose from $2 billion each to $4 billion as 30-year yields approached a near-20-year high. Stan Druckenmiller called it price management; a Wharton economist who served under Bessent calls it plumbing.
Reality
- Evidence52
- Adoption
- Not scored
- Hype gap+25
- Incentives68
- Confidence48
Wharton's agency decay scale tracks how a team slides from testing AI to depending on it
Wharton's Scale of Agency Decay runs experimenting, integrating, relying, depending. The account details the first two, and its usable marker is whether people still overrule a system that contradicts them.
Reality
- Evidence55
- Adoption60
- Hype gap+25
- Incentives55
- Confidence50
A Wharton economist prices the AI buildout's break-even at 2.7x productivity by 2030
Jessica Wachter asked how fast hyperscaler earnings must grow to justify nearly $1.1 trillion of spending through 2027. Her answer is a 2.7x productivity gain by 2030. Anyone forecasting an AI budget now has a number to test.
Reality
- Evidence55
- Adoption62
- Hype gap+12
- Incentives25
- Confidence58
Wharton panel traces how AI is reshaping quantitative finance
Wharton has put the head of Goldman Sachs's research data strategy team on a panel about AI and quantitative finance, and that job title says more about how these desks will hire than the published transcript yet does.
Reality
- Evidence66
- Adoption
- Not scored
- Hype gap+28
- Incentives68
- Confidence54
The Trump administration permanently ends beneficial ownership reporting for U.S. companies
The Patriot Act duties banks actually staff, customer identification and suspicious activity reporting, are intact, and the one filing permanently eliminated carries no published cost. A bank has nothing in the record to book as a saving.
Reality
- Evidence52
- Adoption25
- Hype gap+22
- Incentives65
- Confidence45
A Wharton paper prices the gap between what firms do on the environment and what they say
The working paper puts one standard deviation of greenhushing at 2.29 percentage points of three-month abnormal return, about $238m for its average firm, with conduct measured by third-party sentiment rather than an audit.
Reality
- Evidence52
- Adoption
- Not scored
- Hype gap+16
- Incentives63
- Confidence44
Wharton reads a trillion dollars of AI capex as a bet on tripling productivity
The Wharton model works backwards from the spending to find the productivity gain that would justify it. The answer is 2.7 times per boom, and its authors say the five firms making that bet face solvency risk if it misses.
Reality
- Evidence42
- Adoption25
- Hype gap+30
- Incentives70
- Confidence45
Correcting one assumption turns the EPA's $600 billion rollback benefit into a net cost
Ten economists writing in Science trace the swing between the Biden and Trump vehicle rules to two contested parameters, one of which the 2026 analysis appears to use in two incompatible ways at once.
Reality
- Evidence62
- Adoption22
- Hype gap+12
- Incentives70
- Confidence52
Some employers say they'd rather use AI than hire recent graduates
Two employer surveys say recent graduates arrive unready, while the low-stakes tasks that used to teach them are the first ones handed to AI. The bench for three years out is a spending decision now.
Reality
- Evidence44
- Adoption
- Not scored
- Hype gap+34
- Incentives66
- Confidence41
Earlier coverage
- The $40 trillion problem has a $32.8 trillion denominator
Invest · August 25, 2026 · One report1 publisher