New Report: AI Trust Is Rising in Finance and Business – But So Is the Risk of Getting It Wrong

Survey of 1,000+ decision-makers finds 62% have been negatively impacted by a bad AI output, even as trust in enterprise AI is rising

Key Facts (At-a-Glance):

Launch: AlphaSense issues The 2026 State of AI for Business and Finance report, a global survey of AI decision-makers which found that while trust and adoption of AI tools in the enterprise is rising, inaccuracy of outputs, time spent on manual verification of results, and shadow AI tools and lack of governance continue to cause concerns.

Findings: Productivity gains are tangible - 60% say AI helps them make better decisions and 84% say AI surfaces an insight they would not have found otherwise at least weekly - yet 62% of professionals have been negatively impacted by at least one bad AI output and more than half of professionals are using AI tools not approved by their employer.

Why It Matters: AI decision-makers rank accuracy as the top feature they need in enterprise AI, ahead of data privacy and security, output quality, ease of use, and speed. As AI becomes embedded in new critical workflows, the accuracy of AI’s output, the governance of its use, and return on the investment are areas where leaders can ensure the value of their AI spend.

NEW YORK — September 9, 2026 — AlphaSense, the AI platform redefining market intelligence for the business and financial world, today revealed a new survey of AI decision-makers in enterprises found that trust in AI tools is climbing, but so is the risk of relying on them without adequate verification, exposing a widening gap between AI adoption and AI accountability and trust.

The 2026 State of AI for Business and Finance report is based on a survey conducted by AlphaSense of more than 1,000 leaders and decision-makers with primary responsibility over AI across financial services, corporate, and consulting functions in 11 countries. It finds that AI use has become nearly universal among this group – and that its impact now extends well beyond time savings.

Key findings from the report:

• 84% of respondents say AI surfaces an insight they would not have found otherwise at least weekly
• 64% of investment research professionals say AI has helped them identify a risk earlier
• 60% of respondents say AI use has helped them make better decisions
• Accuracy is the top feature professionals look for in an AI tool, ranked ahead of privacy, speed, and ease of use

Trust has risen alongside that value: trust in enterprise-grade tools increased year over year to 78% versus 72% in 2025, and trust in consumer-grade tools rose to 68% compared to 62% in 2025.

But that value comes with a catch. The survey found that 62% of professionals have been negatively impacted by at least one bad AI output. The specific harms include citing an inaccurate stat (30%), quoting a fabricated source (21%), walking into a meeting with incorrect information (20%), making a wrong decision (20%), and unknowingly misleading a colleague or customer (18%). Investor relations professionals reported the sharpest illustration of these dangers, with 79% having been burned before.

Across AI platforms, 79% of users state they have to verify its output always or often. While human verification of AI output is always the right decision, it also highlights how significant manual effort remains in order to verify claims from AI tools.

"Rising confidence and widespread adoption of AI in enterprises hasn't made the stakes of a wrong answer any less significant," said Sarah Hoffman, Director of AI Thought Leadership at AlphaSense. "It is revealing that the data finds the majority of professionals have been misled by an AI output, and that accuracy is the top feature professionals require today. It is accuracy and verifiability within AI tools that truly reduces manual effort and provides greater return on investment at scale in an organization. As enterprises institute greater governance over their employees’ AI usage and consolidate around AI platforms with trust built in, we’ll begin to see the gap between AI adoption and AI trust narrow.”

A Governance Gap at the Top

The report also uncovers a widening organizational accountability gap. While nearly half of respondents work at organizations that formally mandate or encourage AI use, more than half say they use tools their IT department hasn't approved — a practice known as "shadow AI."

Notably, this behavior is most common among senior leadership. The clearest example: 36% of C-suite leaders report using two or more unapproved AI tools, a higher rate than any other level in the organization.

Further findings reveal:

• Nearly half of respondents state AI is formally mandated or encouraged by their employer, yet close to a third use at least two AI tools their IT department has not approved.
• Regional patterns vary widely: Europe has the strongest formal AI mandates but also the highest shadow-AI use (37%), while APAC shows the lowest trust and urgency but the tightest governance (only 18% run multiple unapproved tools).
• In the U.S., more than half use tools outside IT approval, well above APAC, where 63% report using none at all.
• In Europe, 62% use at least one unapproved tool, the inverse of APAC's compliance rate.
• Shadow AI use is heaviest in investment banking and PE and VC, where 37% run two or more unapproved tools.

Decision-Grade AI Offers Clarity and Accountability

The report demonstrates why decision-grade AI – outputs with defensible citations from source material that a professional can trust and use in work they're personally accountable for – is the standard in the enterprise.

The report concludes, “Trust has risen and AI is enhancing decision-making, but neither has reduced the need to verify the work. Accountability extends beyond checking whether an individual output is accurate; it includes whether organizations can govern how AI is used, and whether leaders can demonstrate that growing AI investments are creating measurable value.”

Download The 2026 State of AI for Business and Finance report here.

Methodology

The findings are based on an unweighted online survey of 1,028 AI decision-makers and leaders in financial services and corporate functions, fielded May 27 to July 30, 2026, across the United States (727), Europe (160), and APAC (141). Respondents span financial services, corporate, and consulting functions, including investment banking, asset management, private equity, investor relations, and retail.

Because the sample was deliberately selected for AI decision responsibility — 63% are the primary decision-maker on AI at their organization and the remainder share that responsibility — results should be read as the views and experiences of AI-engaged business professionals rather than as representative of the entire workforce.

About AlphaSenseAlphaSense is the AI platform redefining market intelligence and workflow orchestration, trusted by thousands of leading organizations to drive faster, more confident decisions in business and finance. The platform combines domain-specific AI with a vast content universe of over 500 million premium business documents — including equity research, earnings calls, expert interviews, filings, news, and internal proprietary content. Purpose-built for speed, accuracy, and enterprise-grade security, AlphaSense helps teams extract critical insights, uncover market-moving trends, and automate complex workflows with high-quality outputs. With AI solutions like Generative Search, Generative Grid, and Deep Research, AlphaSense delivers the clarity and depth professionals need to navigate complexity and obtain accurate, real-time information quickly. For more information, visit www.alpha-sense.com.

Media Contact

Pete Daly for AlphaSense

Email: media@alpha-sense.com