Artificial intelligence has entered the workplace.

Employees are using AI to draft emails, summarise meetings, research customers, analyse information, create marketing content and solve technical problems.

However, widespread use does not necessarily mean successful implementation.

Many businesses are using AI to complete individual tasks slightly faster while leaving the wider business process unchanged. Employees still copy information between systems, chase customers manually, prepare repetitive reports and depend on disconnected software.

The real opportunity is not simply using an AI chatbot.

It is connecting AI, business automation and custom software so that complete processes can operate faster, more accurately and with less manual intervention.

How Are People Using AI in 2026?

A major 2026 study examined 12,637 practical AI use cases gathered from a wider database of almost 50,000 records.

The research found that generative AI is being used across a growing range of personal, technical and professional activities. It also found that the way people use AI is becoming broader and more sophisticated, although human judgement remains essential.

The study was global rather than specifically focused on UK SMEs. However, its findings reveal several important patterns that apply directly to small and medium-sized businesses.

People are increasingly using AI to:

• Troubleshoot problems
• Draft and edit documents
• Support business decisions
• Analyse information
• Create content
• Assist with software development
• Organise work
• Automate simple actions
• Support entrepreneurs and start-ups
• Convert notes into useful outputs

The critical issue is how these activities are being introduced into businesses.

In many cases, employees are adopting AI independently rather than through an organised company-wide strategy.

AI Use at Work Is Often Employee-Led

The research found that 63 of the top 100 AI use cases were either directly connected to work or applicable in both professional and personal settings.

Common workplace uses included decision support, career assistance, entrepreneurship, technical troubleshooting and using AI as an informal work assistant.

This demonstrates that AI adoption is not waiting for formal corporate programmes.

Employees are already experimenting with tools such as ChatGPT, Gemini, Copilot and specialist AI platforms.

They may be using them to:

• Improve an important email
• Summarise a long document
• Prepare for a customer meeting
• Write a job description
• Analyse a spreadsheet
• Create a presentation
• Develop a marketing campaign
• Research a competitor
• Draft a proposal
• Explain a technical problem

These uses can save time, but they are usually isolated.

The employee receives a faster output, but the rest of the process remains manual.

The Difference Between Using AI and Implementing AI

There is a significant difference between an employee using AI and a business implementing AI.

Using AI

An employee opens an AI tool, enters instructions and receives an output.

For example:

1. A sales enquiry arrives.
2. The salesperson copies the customer’s information.
3. They enter it into an AI platform.
4. AI drafts a response.
5. The salesperson checks the response.
6. They copy it into their email.
7. They update the CRM.
8. They set a reminder to follow up.

AI has helped with one step: writing the email.

Implementing AI

A properly connected system could:

1. Detect the new enquiry.
2. Categorise the customer’s request.
3. retrieve relevant information from the CRM.
4. Draft a personalised response.
5. Send it for approval or automatically respond.
6. Update the customer record.
7. Assign the opportunity to the correct salesperson.
8. Schedule the follow-up.
9. Notify management if no action is taken.
10. Record the outcome for reporting.

This is more than AI assistance.

It is business process automation powered by AI.

Why Most Businesses Are Only Seeing Small Improvements

The research suggests that workplace AI use remains fragmented.

Companies may purchase AI licences, arrange training sessions and encourage experimentation. At the same time, employees may face unclear policies, restricted tools, concerns about confidentiality or uncertainty about whether they are allowed to use AI.

This has contributed to “shadow AI”: employees using AI without openly discussing it with management.

Some employees reported completing work substantially faster while keeping their AI use private because the company had not formally approved or understood the technology.

The result is an inconsistent approach.

One employee may use AI effectively. Another may avoid it completely. A third may enter sensitive business information into an unsuitable public tool.

The company gains some individual productivity, but it does not develop a reliable operating system for AI.

Where Businesses Are Currently Getting Value From AI

Most current business uses fall into three categories.

1. Efficiency

AI can accelerate work that already exists.

Common examples include:

• Summarising meetings
• Preparing first drafts
• Producing internal documents
• Analysing customer feedback
• Explaining financial information
• Creating reports
• Processing recruitment information
• Categorising enquiries
• Extracting information from documents
• Reducing repetitive administration

These applications can save employees time and improve consistency.

However, efficiency gains remain limited when the AI tool is disconnected from the company’s other systems.

2. Growth

AI can support revenue-generating activities such as sales, marketing and customer engagement.

Examples include:

• Personalising sales emails
• Identifying promising leads
• Producing campaign variations
• Creating customer-specific proposals
• Analysing sales conversations
• Recommending follow-up actions
• Testing different marketing messages
• Answering customer enquiries outside normal working hours
• Re-engaging previous customers
• Identifying cross-selling opportunities

One example in the research described AI-assisted email personalisation and continuous testing producing a reported 20% to 30% improvement for clients. However, the study noted that clearly measured returns remained relatively uncommon.

This is an important lesson for business owners.

Claims about AI are easy to make. Measurable commercial results are more valuable.

3. Business Transformation

Transformation happens when a company changes how work is completed rather than simply making an existing task faster.

This could involve:

• Automating an entire customer onboarding process
• Building a system that produces quotations automatically
• Giving managers real-time operational information
• Connecting sales, customer service and finance systems
• Creating an AI agent that completes defined actions
• Replacing several spreadsheets with one internal application
• Allowing customers to complete processes without staff intervention
• Redesigning a service around faster digital delivery

The study found that these more transformational implementations were still relatively early, but that SMEs appeared among the businesses benefiting from them.

Small businesses can sometimes move faster because they have fewer approval layers, less legacy technology and direct access to decision-makers.

AI Agents Are Moving From Conversation to Action

Traditional generative AI mainly responds to questions and produces content.

AI agents can go further.

An AI agent may be able to:

• Read incoming information
• Make a defined assessment
• Select an appropriate next step
• Update a business system
• Create a task
• Send a message
• Schedule an appointment
• Request missing information
• Escalate a problem
• Continue a process without constant human instructions

Autonomous agentic operations reached sixth place among the study’s leading AI use cases in 2026.

Many examples were still experimental and relatively small. One common pattern involved converting voice notes or written notes into organised information and routing them into the correct system or next action.

For SMEs, this could mean turning a business owner’s voice note into:

• A task for an employee
• A customer follow-up
• A calendar appointment
• A CRM update
• A quotation request
• A maintenance ticket
• A management note

The value comes from removing the manual steps between the original instruction and the completed action.

Vibe Coding Is Expanding Access to Custom Software

Vibe coding refers to using natural-language instructions to create or modify software.

Instead of writing every line of code manually, a user describes what the application should do and AI assists with producing the code.

The research identified vibe coding as another major emerging use case in 2026.

This does not mean that every business owner should attempt to build critical software without technical oversight.

Business systems still require:

• Clear requirements
• Secure architecture
• Reliable integrations
• Data protection
• Testing
• Error handling
• User permissions
• Ongoing maintenance
• Human accountability

However, AI-assisted development can reduce the time and cost required to create certain internal tools, prototypes and custom applications.

This makes custom software for SMEs more accessible than it was previously.

Practical AI Opportunities for Small Businesses

AI implementation should begin with a business problem, not a technology.

The following areas frequently contain opportunities.

Sales

AI and automation can:

• Qualify new enquiries
• Assign leads
• Prepare follow-up messages
• Update CRM records
• Identify inactive opportunities
• Summarise sales calls
• Recommend the next action
• Produce quotations
• Notify managers about missed follow-ups

Customer Service

Businesses can use AI to:

• Answer common questions
• Categorise support requests
• Retrieve relevant account information
• Draft accurate responses
• Route complex cases
• Identify frustrated customers
• Follow up after resolution
• Provide service outside standard hours

Finance and Administration

Potential applications include:

• Extracting data from invoices
• Matching receipts
• Identifying overdue payments
• Preparing payment reminders
• Categorising documents
• Updating financial systems
• Producing cash-flow summaries
• Highlighting unusual transactions

Marketing

AI can help businesses:

• Research customer questions
• Plan content
• Repurpose existing material
• Create campaign variations
• Personalise messages
• Analyse campaign results
• Identify successful topics
• Maintain a consistent publishing schedule

Operations

Operational opportunities include:

• Staff scheduling
• Stock monitoring
• Supplier communication
• Maintenance reporting
• Compliance reminders
• Management dashboards
• Task assignment
• Document generation
• Process tracking

The best opportunity is usually a process that is repetitive, time-consuming, measurable and currently dependent on manual work.

The Risk of Poor AI Implementation

AI can produce impressive-looking outputs even when the underlying information is wrong.

Businesses should not allow AI to make uncontrolled decisions simply because it can produce a confident answer.

Poor implementation can create:

• Incorrect customer communication
• Confidentiality breaches
• Inaccurate reporting
• Unapproved commitments
• Biased decisions
• Duplicate work
• Unclear accountability
• Dependence on unreliable outputs
• New software costs without measurable savings

AI should therefore be introduced with clear boundaries.

Businesses need to decide:

• Which tools employees may use
• What information may be entered
• Which outputs require human approval
• Which actions AI can complete automatically
• Who is responsible when something goes wrong
• How performance will be measured

The objective is controlled automation, not uncontrolled experimentation.

A Practical AI Implementation Framework

Businesses can use the following process to identify valuable AI opportunities.

Step 1: Map the Current Process

Document how the work happens today.

Identify:

• Who completes each step
• Which systems are involved
• Where information is copied
• Where delays occur
• Which decisions require judgement
• Which actions follow predictable rules

Step 2: Calculate the Cost

Measure:

• Employee hours
• Management time
• Error rates
• Missed sales
• Customer waiting time
• Software costs
• Repeated corrections

This establishes the potential value of improving the process.

Step 3: Separate Human and Automated Work

Not every task should be automated.

Human involvement remains important for:

• Strategic decisions
• Complex negotiations
• Sensitive customer situations
• Creative direction
• Legal accountability
• Relationship management
• Final approval of high-risk actions

AI and automation are better suited to repetitive, structured and information-heavy work.

Step 4: Choose the Right Technology

The solution may involve:

• An existing AI platform
• Workflow automation
• CRM automation
• An AI agent
• System integration
• A custom internal application
• Custom AI software
• A combination of several approaches

The most advanced option is not automatically the best one.

The correct solution is the one that produces a reliable commercial result.

Step 5: Start With One Measurable Process

Do not attempt to automate the entire business at once.

Choose one process where success can be measured through:

• Hours saved
• Faster response times
• Lower operating costs
• Fewer mistakes
• More completed follow-ups
• Increased sales
• Improved customer satisfaction
• Reduced dependence on the owner

Step 6: Test, Measure and Expand

Compare the new process with the original baseline.

Expand the implementation only when it produces a clear improvement.

The Bigger AI Opportunity for SMEs

Many businesses currently use AI as a faster writing or research tool.

That is useful, but it represents only the first stage of adoption.

The greater opportunity is creating a business where:

• Enquiries are handled immediately
• Customer information moves automatically
• Repetitive administration is removed
• Managers receive accurate information without preparing reports manually
• Sales opportunities are followed up consistently
• Employees spend more time on valuable work
• The company can grow without increasing administration at the same rate

This requires more than prompts.

It requires understanding the business process, selecting the right technology and building systems that work together.

How Nuvaleo AI Helps Businesses Implement AI

Nuvaleo AI helps small and medium-sized businesses use artificial intelligence, business automation and custom software to improve how they operate.

We begin by examining the business itself—not by selling a predetermined AI tool.

We identify:

• Repetitive manual work
• Disconnected software
• Time-consuming administration
• Missed sales opportunities
• Slow customer-response processes
• Reporting bottlenecks
• Tasks that depend unnecessarily on the business owner
• Areas where automation can create measurable value

We then determine whether the right solution is:

• AI implementation
• Workflow automation
• An AI agent
• Software integration
• A custom internal application
• Purpose-built business software

We show business owners what their business could do automatically, then build it for them.

Frequently Asked Questions

How are businesses using AI in 2026?

Businesses are using AI for writing, research, customer communication, data analysis, software development, meeting summaries, sales support and repetitive administrative work. More advanced companies are beginning to use AI agents and connected automation to complete actions across business systems.

Can small businesses benefit from AI?

Yes. Small businesses can often implement AI quickly because their processes are less complicated and decision-makers are more accessible. The strongest opportunities normally involve repetitive administration, customer communication, sales follow-up and reporting.

What is business automation?

Business automation uses software to complete repeatable tasks or move information between systems with less manual involvement. AI can strengthen automation by interpreting unstructured information, drafting content and selecting appropriate actions.

What is an AI agent for business?

An AI agent is a system that can assess information and complete defined actions. For example, it could categorise an enquiry, retrieve customer information, prepare a response, update the CRM and schedule a follow-up.

Does every business need custom AI software?

No. Many problems can be solved with existing platforms or straightforward workflow automation. Custom software becomes valuable when existing products do not match the company’s process, several systems must be connected or the workflow provides a competitive advantage.

How should a business start using AI?

Start by identifying one expensive, repetitive or slow process. Measure its current performance, determine which parts require human judgement and test a focused solution with a measurable objective.

Turn AI Into a Business System

Using AI for isolated tasks may save a few minutes.

Connecting AI to your processes can change how the entire business operates.

Nuvaleo AI helps businesses find where time is being lost, identify where manual work can be removed and build practical solutions around measurable commercial outcomes.

Book a Business Profit and Efficiency Assessment to discover what your business could do automatically.