Skip to content
Fast Company Executive Board

The Fast Company Executive Board is a private, fee-based network of influential leaders, experts, executives, and entrepreneurs who share their insights with our audience.

Which AI solution is right for you?

Know the differences and ask the right questions before you decide.

[Photo: Getty Images]

Profile photo of Jamie Timm

Chief Delivery Officer · Telus Digital Solutions

Business leaders often face pressure to adopt emerging technologies quickly, sometimes before fully assessing the risks and potential for long-term value. Especially so with AI, as compounding deployment urgency over the past five years created both transformative opportunities and far-reaching consequences.

In that time, AI progressed from specialized predictive systems to large foundation models, and from there into enterprise-scaled generative AI (GenAI) applications, copilots, and autonomous and agentic agents.

In customer experience (CX), AI is increasingly deployed across every touchpoint, from initial customer inquiries to complex service resolution. Enhancing CX while managing costs and improving efficiency raises the stakes for selecting the right AI solution.

Notably, the market’s rapid AI product evolution and corresponding product influx has resulted in conflation of very different AI solutions, increasing the difficulty of finding the right technology to address organizational needs.

This could be one reason why 42% of companies abandon the majority of their AI initiatives before reaching production, up from just 17% in 2024, according to S&P Global’s Voice of the Enterprise: AI & Machine Learning, Use Cases 2025.

So how can leaders move beyond the confusion? Assess the role each solution can play. Then, engage in a quick, initial series of questions to assess organizational needs. These two steps can get your organization confidently started down the right AI path.

Understand AI solution differences and uses

Clients consistently face difficulty in distinguishing between the various types of AI, and understanding which best align with their business needs and goals. Each AI type has its place in the CX ecosystem and can work with others as complementary solutions, but organizations must understand their individual capabilities to leverage their strengths.

For customer answers: Chatbots

  • Respond to customer inquiries through text or voice.
  • Handle straightforward questions (store hours, account balances, basic troubleshooting).
  • Understand natural language but operate within limited scope.

For customer support: Assistants (co-pilots)

  • Support human agents by providing information and tailored response suggestions.
  • Streamline workflows and save agent time.
  • Require human oversight and cannot act independently.

For customer-focused actions: Three AI choices

Automations

  • Run predefined, rule-based processes.
  • Execute repetitive tasks (order processing, FAQ responses, product information delivery).
  • Highly efficient and reliable but lack flexibility and decision-making capabilities.

Agents

  • Handle customer interactions and workflows autonomously within established guardrails.
  • Engage in dynamic conversations and make contextual decisions.
  • Independently manage complete customer journeys from start to finish.
  • Take autonomous actions (processing returns, updating accounts).
  • Escalate to humans when situations exceed capabilities or require specialized expertise.

Agentic AI

  • A subset of the broader category of autonomous agents.
  • Capable of multi-step reasoning, adapting to context, taking initiative, and chaining actions across tools or environments.
  • According to Gartner, by 2029, agentic AI will autonomously resolve 80% of common customer service issues without human intervention, leading to a 30% reduction in operational costs.

Why the right AI matters

As you might expect, misapplication of the various AI types can result in underperformance, lost investment, and erosion of trust, both within an organization and with customers. A mismatch can result in several potential scenarios, including:

  • Poor customer experience if a chatbot can’t manage complex queries.
  • Operational inefficiencies from deploying overpowered solutions for simple tasks.
  • Underutilized technology sitting idle because it doesn’t match actual business needs.

Moreover, a mismatch can lead to employee change fatigue. Employees lose confidence in adoption if the chosen tool doesn’t meet their needs or creates more work, leading to resistance and ultimately failed implementations.

Match AI to business needs

To determine the best AI solution for your organization, the following categories and questions may be helpful.

1. Determine autonomy needs

  • Do you need an independent AI solution? AI agents can deliver autonomy for end-to-end processes.
  • Do you need human oversight? Chatbots or assistants with escalation pathways are better options when nuanced decision making and human judgment is critical.
  • What are your channel preferences?[1] [2]  Identify where your customers engage most (web, mobile, voice, messaging) and ensure your AI solution integrates seamlessly.
  • What are your team’s skill gaps and efficiency levels? Use current gaps and efficiencies to gauge AI’s application.

2. Balance task complexity

  • Are your tasks simple and repetitive? Automation is ideal for data entry, order processing, and routine tasks.
  • Do you need to support human agents? Assistants or copilots excel at expediting and streamlining workflows. Incorporate human-in-the-loop or human validation during testing and implementation to ensure AI recommendations are accurate and align with customer needs.
  • Do you require complex problem-solving? AI agents can handle sophisticated, multi-step processes.
  • What is your technical readiness? Confirm the existence of model context protocol (MCP) servers for applications you need to interact with and ensure appropriate application programming interfaces (APIs) are available for seamless integration. 

3. Weigh your risk tolerance

  • What’s your inaccuracy tolerance? AI chatbots interact with customers, and can make mistakes. 
  • How critical are data security, privacy protection and ethical applications? Align autonomy with your risk appetite.
  • Have you implemented security testing and red-teaming? As of 2025, 67% of enterprises report a shortage of cybersecurity professionals to address GenAI adoption-related safety concerns and security vulnerabilities. Automated solutions can help organizations identify pre-deployment AI vulnerabilities by running thousands of adversarial attacks in minutes.

The path to successful AI implementation

Successful AI implementations match the appropriate tool to the specific task at hand. A trusted partner can help you understand each technology’s capabilities and constraints, and assess your organization’s needs and risk tolerance. Companies can implement AI more strategically, avoid costly missteps, and achieve measurable business outcomes that drive long-term success.

Jamie Timm is the global SVP, service delivery and operations[3]  at TELUS Digital Solutions.