August 28, 2026
This is the third in a series of general education articles of relevant topics that we think you may find of interest. If nothing else, this insight may help you look and sound more knowledgeable with your golf foursome, at pickleball, or at your local watering hole.
Agentic AI is a type of artificial intelligence designed to work toward a goal with some independence. Rather than answering one question and stopping, an AI agent can assess an objective, break it into steps, choose the right tools or data sources, take action, review the result, and decide what to do next. Put simply, generative AI is usually focused on producing an answer or piece of content. Agentic AI is focused on getting something done.
A simple example helps show the difference. You might ask a chatbot, “What flights are available to New York?” and receive a list of options. With agentic AI, the request could be broader: “Find the best flight for my business trip, compare it with my calendar and travel policy, prepare an itinerary, and ask me before booking.” In the first case, the system provides information. In the second, it plans, searches, compares, checks constraints, and prepares an action. That movement from answering to acting is what makes agentic AI significant.
Most agentic systems start with a large language model, then add capabilities that let the system do more than generate text. Planning allows the agent to turn a broad goal into smaller tasks. Tool use lets it search the web, query a database, call an API (application programming interface which is a set of rules), write code, update a spreadsheet, send a message, or work inside business software. Memory and context help it track what it has already tried, what information it has gathered, and what rules or limits apply. Feedback loops allow the agent to check whether each step is working and adjust course when needed.
That is why agentic AI sits somewhere between basic automation and a standard chatbot. Traditional automation follows fixed instructions: if this happens, do that. A chatbot usually responds to a single request. An agent is more flexible than a scripted workflow, but it should still operate within clear permissions, rules, and human oversight.
The most common uses today are in work that are digital, repetitive, multi-step, and spread across several systems. In software development, coding agents can draft code, run tests, find bugs, suggest fixes, and document changes. They do not replace skilled developers, but they can make routine programming and debugging faster.
Customer service is another natural fit. Agents can sort incoming requests, pull account details, draft responses, escalate unusual issues, and update records. In sales and marketing, they can research prospects, personalize outreach, review campaign performance, and recommend follow-up steps. In finance and operations, they can reconcile invoices, flag unusual transactions, prepare reports, track supply-chain problems, and route approvals. For research and analysis, they can gather information from different sources, summarize findings, compare options, and create a first draft of a decision memo.
Personal productivity may be one of the most visible uses. An agent could help manage email, schedule meetings, prepare briefing materials, organize documents, monitor deadlines, or coordinate travel. Healthcare, cybersecurity, logistics, legal work, finance, and education are also likely to adopt these tools, although those higher-stakes areas need stronger guardrails because mistakes can carry serious consequences.
Productivity and efficiency. The clearest benefit is speed. Agents can move across documents, databases, calendars, email, software tools, and websites without a person manually transferring each piece of information. That can reduce administrative work and give people more time for judgment, relationships, creativity, and strategy.
Scalability. A well-designed agent can handle large volumes of routine work without needing the same level of human attention each time. It can monitor systems around the clock, respond quickly, and apply the same process consistently. For businesses, which can lower the cost of serving customers, processing information, or managing internal workflows.
Better use of software. Many organizations rely on applications that do not always work smoothly together. Agentic AI can help connect those systems by retrieving information from one place, updating another, and coordinating steps that used to require manual effort. This is one reason agentic AI is viewed as an important shift in enterprise software: users may be able to state what they want done instead of clicking through every step themselves.
Accessibility of expertise. Agents can make specialized help easier to access. A small business may not have a full-time analyst, programmer, compliance assistant, or operations coordinator, but an AI agent could handle parts of those roles. Used carefully, this can broaden access to skills that were once too expensive or unavailable.
Adaptability. Agents are also more adaptable than rigid automation. If a data source is missing, a website changes, or a first answer is incomplete, the agent can try another path. That makes it useful in real-world workflows, where every situation does not follow a perfect script.
Error propagation. The biggest weakness is that agents make decisions across several steps. A small misunderstanding early in the process can grow into a larger failure later. A chatbot that gives the wrong answer may be inconvenient. An agent that sends the wrong message, changes the wrong file, or books the wrong transaction can create much bigger problems.
Security and privacy. To be useful, agents often need access to sensitive systems and data. That access creates risk if permissions are too broad, if the agent follows malicious instructions, or if confidential information is exposed. Prompt injection, unauthorized actions, data leakage, and misuse of credentials are among the most serious concerns.
Accountability and transparency. The more independently agents operate, the more important it becomes to know why they took a particular action. That matters in regulated industries, client-facing work, healthcare, finance, legal services, and any setting where decisions need to be reviewed. Organizations need clear logs, approval rules, and defined responsibility for what agents do.
Overreliance. Convenience can also make people less careful. If an agent usually works well, users may stop checking its output closely. That is risky because AI can sound confident even when it is wrong. Human review is still essential for high-stakes decisions, unusual cases, sensitive communications, financial commitments, and anything involving safety or legal responsibility.
Workforce disruption. Agentic AI may automate not only isolated tasks, but also larger workflows. Some jobs may shrink, disappear, or change significantly. At the same time, new roles will likely emerge around supervising agents, designing workflows, managing data, setting policy, and handling exceptions. The technology may improve productivity overall, but the transition could be difficult for workers and organizations that are not prepared.
Agentic AI is best understood as the next step beyond generative AI. It moves AI from producing content to pursuing goals and taking action. Its strongest uses are likely to be bounded, repeatable, digital workflows where the agent has clear instructions, limited permissions, and human oversight. The upside is meaningful: faster execution, lower costs, better use of software, and broader access to expertise. The downside is just as real: errors can build on one another, security risks increase, accountability becomes harder, and workers may face disruption.
The most realistic future is not fully autonomous AI replacing people overnight. It is people supervising increasingly capable digital agents. The organizations that benefit most will be the ones that start with narrow use cases, limit access, keep audit trails, require approval for high-risk actions, and train users to verify important outputs. Agentic AI could become a major productivity layer in the economy, but its success will depend less on raw intelligence and more on trust, governance, security, and practical human control.
Landsberg Bennett is a group comprised of investment professionals registered with Hightower Advisors, LLC, an SEC registered investment adviser. Some investment professionals may also be registered with Hightower Securities, LLC (member FINRA and SIPC). Advisory services are offered through Hightower Advisors, LLC. Securities are offered through Hightower Securities, LLC.
This is not an offer to buy or sell securities, nor should anything contained herein be construed as a recommendation or advice of any kind. Consult with an appropriately credentialed professional before making any financial, investment, tax or legal decision. No investment process is free of risk, and there is no guarantee that any investment process or investment opportunities will be profitable or suitable for all investors. Past performance is neither indicative nor a guarantee of future results. You cannot invest directly in an index.
These materials were created for informational purposes only; the opinions and positions stated are those of the author(s) and are not necessarily the official opinion or position of Hightower Advisors, LLC or its affiliates (“Hightower”). Any examples used are for illustrative purposes only and based on generic assumptions. All data or other information referenced is from sources believed to be reliable but not independently verified. Information provided is as of the date referenced and is subject to change without notice. Hightower assumes no liability for any action made or taken in reliance on or relating in any way to this information. Hightower makes no representations or warranties, express or implied, as to the accuracy or completeness of the information, for statements or errors or omissions, or results obtained from the use of this information. References to any person, organization, or the inclusion of external hyperlinks does not constitute endorsement (or guarantee of accuracy or safety) by Hightower of any such person, organization or linked website or the information, products or services contained therein.
Click here for definitions of and disclosures specific to commonly used terms.