Investment banking runs on two things: information advantage
and trust. AI is now reshaping how both get built, across research, diligence,
and the deal cycle itself.
Adoption is moving fast at the top. In banking, generative
and agentic AI now make up 70% of publicly announced AI implementations, up
from 54% a year earlier, according to Evident's Q4 2025 tracking.[1] The banks
Evident studies are the largest in the world, but the use cases scale straight
down to a boutique advisory shop or middle-market bank of 10 to 300 people.
The upside is real, and so is the risk, because everything a
banker touches is confidential and often market-moving. Here are 10 of the most
practical ways banks and advisory firms are using AI heading into 2026 and
2027, with a note on how to adopt each one without exposing material non-public information (MNPI) or
client data.
1. Market and Company Research
Analysts spend hours pulling together filings, news, broker
notes, and transcripts before they can even start thinking. AI collapses that
gathering step, reading across sources and drafting a first synthesis in
minutes.
That turns a junior analyst's overnight research grind into
a same-morning starting point, so the team spends its time on judgment instead
of assembly.
Where the right IT partner helps: Research
pulls in licensed data and sometimes confidential material. Use approved, firm-controlled tools with clear rules on what data they may touch, not free consumer apps on
personal logins.
2. Pitch Books and Client Materials
First drafts of pitch books, market updates, and
confidential information memoranda are heavy on structure and repetition, which
is exactly what AI drafts well. It builds the skeleton, populates standard
sections, and formats to the house template.
Bankers still own the story and the numbers, but they start
from a draft instead of a blank page, which is where the late nights usually
go.
Where the right IT partner helps: Client materials carry confidential deal
information from the first draft. A partner can wire these tools into approved
systems with access limited to the deal team so nothing leaks to a shared
model.
3. Financial Modeling Support
AI can help populate and quality-check models: pulling
historicals into a template, flagging broken links or formula errors, and
running sensitivity scenarios on command.
The discipline that keeps this safe is human ownership of
every assumption and output. AI checks the plumbing; the banker signs off on
the analysis.
Where the right IT partner helps: Models hold deal assumptions you cannot
afford to leak or lose. Good IT management keeps them in controlled,
well-backed-up systems with strict access, not scattered across personal
drives.
4. Due Diligence and Data-Room Review
Diligence is a reading problem at scale. AI reads thousands
of data-room documents, extracts key terms, and flags the change-of-control
clauses, off-balance-sheet items, and anomalies a team needs to look at.
That shortens the slow, expensive first pass and lets senior
bankers focus on the findings that actually move a deal.
Where the right IT partner helps: Data rooms are among the most sensitive
environments in the business. Keep review inside approved, access-controlled
tools, and document who touched what for the record.
5. Comparable Company and Precedent Transaction Analysis
Building comps and precedent sets is repetitive and
error-prone by hand. AI pulls candidates, normalizes metrics, and drafts the
summary tables and rationale, so analysts refine rather than assemble.
The result is faster turnarounds on valuation work and fewer
copy-paste mistakes carried through a live process.
Where the right IT partner helps: Valuation work often blends licensed data
with confidential inputs. A partner can standardize the firm on tools that
respect both data-license terms and confidentiality.
6. Earnings and Transcript Summarization
Quarterly calls, investor days, and management meetings
generate more transcript than anyone can read. AI turns hours of calls into a
tight brief with the quotes and numbers that matter.
For coverage teams tracking dozens of names, that is the
difference between staying current and falling behind.
Where the right IT partner helps: This is a low-risk place to start because
much of the input is public. A partner can still make sure the outputs land in
firm systems, not personal note apps.
7. Deal Sourcing and Relationship Intelligence
AI mines your CRM, news, and filings to surface which
relationships are heating up, which companies fit a buyer's mandate, and where
a banker should spend the next call.
For a boutique competing on relationships, this turns
scattered notes and inboxes into a coverage edge.
Where the right IT partner helps: Relationship data is a core asset and a
privacy responsibility. A partner can connect these tools to your CRM
correctly, with permissions that match who should see what.
8. Engineering and Developer Augmentation
For firms that build their own tools, AI coding assistants
are becoming standard. In banking, developer augmentation use cases have grown
8x since early 2024, and banks report average productivity gains of 10% to 20%.[1]
Even a small internal tech team can ship integrations and
internal tools faster with the same assistants the largest banks now issue by
default.
Where the right IT partner helps: Developer tools reach into source code and
systems. A partner can set the access controls and review steps so speed
doesn't come at the cost of security.
9. Agentic Automation of Back-Office Workflows
The frontier is agentic AI that completes multi-step
workflows, not just answers questions. Banks are already deploying it for
processes like know-your-client onboarding, source-of-wealth reporting, and
fraud claims.[1]
For a smaller firm, the same pattern automates onboarding
paperwork, compliance checklists, and other repeatable back-office work that
ties up staff today.
Where the right IT partner helps: Agents that take actions need tight
guardrails and human checkpoints. A partner can design the controls so an agent
escalates the important decisions instead of acting alone.
10. Compliance and Communications Oversight
AI is increasingly used to review communications and flag
conduct or recordkeeping issues before they become an enforcement problem. For
a regulated firm, that is a control, not a convenience.
Used well, it helps compliance cover more ground with the
same headcount, and catches the off-channel message or risky phrase a human
sweep would miss.
Where the right IT partner helps: Surveillance tools touch sensitive
communications and must themselves be governed. A partner can help you deploy
them with the right retention, access, and audit trail.
The Bottom Line
AI is moving from experiment to infrastructure across
banking, and the vendor field is still shifting. One provider's share of public
banking use cases fell from 35% to 27% in a year as more tools entered the
market.[2] For a smaller firm, that churn is a reason to standardize on a
governed set of tools rather than let every desk pick its own.
That governance gap is where firms get hurt. IBM found 63%
of organizations have no AI governance policy, and high use of unsanctioned
shadow AI added an average of 670,000 dollars to the cost of a breach.[3] For a
bank, the material at risk isn't just data, it's MNPI and client
confidentiality, so an ungoverned tool is a compliance and reputational
problem, not only a security one.
Framework IT helps banks and advisory firms capture the
upside of AI without the downside. That means governance first, a written usage
policy, a review step before new tools go live, and clear rules on what
confidential data a tool may touch, then enablement through the Managed Framework AI adoption program, where staff get
hands-on training and a central Framework AI Resources Hub so they use the
approved tools well. Framework IT is a Chicago-based managed IT services firm
founded in 2008, with a team of more than 40 professionals, most of them
engineers who live in the Chicagoland area. We help investment banks, advisory
shops, and other financial and professional services firms with IT support,
strategy, and security, and with putting structure around AI so it can be used
safely.
Schedule a
conversation with our team to see what safe, high-value AI adoption can look
like for your firm: frameworkit.com/discoverycall
About the Author
Adam Barney is President and Managing Partner of Framework
IT, a Chicago-based managed IT services firm he's helped lead for more than 15
years. He and his team of 40+ professionals specialize in IT support, strategy,
and cybersecurity for small and mid-sized businesses. Adam's insights on
business technology have been featured in the Harvard Business Review, the
Washington Post, and Fox 32 Chicago.
Citations
Every
statistic above is sourced to a live page that states it. Verify links are live
before publishing.
[1] Generative and agentic AI now represent 70% of
publicly announced banking AI implementations, up from 54% a year earlier;
engineer and developer augmentation use cases have grown 8x since February
2024, with banks reporting average productivity gains of 10% to 20%; banks are
deploying agentic AI for processes such as source-of-wealth reporting and fraud
claims. Evident, AI Use Case Trends in Banking, Q4 2025.
https://evidentinsights.com/insights/use-case-trends-q4-2025
[2] One AI provider's share of public banking use
cases declined from 35% to 27% year on year as more vendors entered the market.
Evident, AI Use Case Trends in Banking, Q4 2025.
https://evidentinsights.com/insights/use-case-trends-q4-2025
[3] 63% of organizations have no AI governance
policy; a high level of shadow AI added an average of 670,000 dollars to breach
costs; 97% of organizations with an AI-related incident lacked proper AI access
controls. IBM 2025 Cost of a Data Breach Report. https://www.ibm.com/think/x-force/2025-cost-of-a-data-breach-navigating-ai
