You are right in that many applications have APIs for directly “talking” to those applications. Even Automation Anywhere has its own API for automating control room functionality.

And that’s where I think you have missed part of the picture. If you want to take data from one application and migrate it to another application where BOTH applications have APIs, you can’t just tie them together. You need something to connect them. Automation Anywhere can do this. 

Yes, you can make a PowerShell or Python call to do the same API calls, but where Automation Anywhere shines is being able to not only make those connections, but then perform other business logic with them, such as connecting to a database and merging the data there, or taking that data and using UI manipulation to put that data into an application that doesn’t have an API, or being able to enter data into a mainframe application via Terminal Emulation.

That’s the strength of Automation Anywhere: I can connect X to Y to Z even if they were never designed to be connected. Add in Document Automation and its native AI document extraction capabilities and you’ve got another source of data that can be automated.

I’d also like to weigh in as a long time AutoIt user, someone who has contributed to the language, written his share of UDFs, and still moderates the forums. I have a lot of love for the language, and always will - but comparing AutoIt to Automation Anywhere is misinformed at best.

AutoIt is based off the Win API, so you are limited in what it can do. We have for years tried to push it beyond those limitations, but it is what it is. Your colleague may be using Automation Anywhere for things that AutoIt can do (finding and interacting with elements in web pages or thick applications) but not only is that not all Automation Anywhere can do, I would say it isn’t even the main thrust any longer.

From API connections which are 100 times easier to work with in AA than AutoIt (check out Connector Builder, so awesome), to AI Agents, to integrations with so many backend systems and applications (Salesforce, M365, Google, SAP, ServiceNow, the list goes on) that it is mind boggling, to tasks that run solely in the cloud with no local infrastructure needed, not to mention all the items Aaron mentions above: I would say if your colleague is only using it to point and click, they are missing out. 

Automation Anywhere remains highly relevant in 2025, especially in scenarios where APIs aren't available, or where legacy systems and applications must be integrated. While APIs are powerful for direct communication between applications, they often require meticulous programming, and some systems—like mainframes or proprietary applications—might lack API support altogether. This is where Automation Anywhere provides a distinct advantage by bridging the gap through GUI-based automation or even terminal emulation.

Additionally, the platform enables business users without deep programming knowledge to create automation workflows. This democratization of automation makes it accessible to a broader audience, fostering innovation within teams that may not have dedicated developers.

Beyond connecting systems, Automation Anywhere incorporates advanced capabilities like AI-driven document processing (e.g., extracting key data fields from invoices) and enables end-to-end process orchestration, which includes error handling and validations. This makes it more than just a tool for automation—it becomes a comprehensive business process solution.

Furthermore, while APIs are excellent for efficiency, they often focus on point solutions and require bespoke development for larger processes. RPA tools like Automation Anywhere help organizations scale automation projects by offering centralized management, monitoring, and governance, ensuring that processes remain consistent and resilient.

This isn't to say APIs and RPA tools are mutually exclusive—they can complement one another. For example, APIs can be embedded within RPA workflows for streamlined operations, enhancing the power of automation solutions.

While RPA tools can integrate with both new and legacy systems, their true advantage lies primarily in working with legacy systems lacking modern integration capabilities.

For environments with only new systems that already provide robust API support, investing in RPA tools may be unnecessary and cost-inefficient. In such cases, you could implement more economical API integration solutions without incurring:

• RPA licensing costs
• RPA infrastructure expenses
• Additional maintenance overhead

Although RPA platforms do offer API execution capabilities, they're typically not the most cost-effective option when your needs are strictly limited to API integrations. Purpose-built API integration tools, iPaaS (Integration Platform as a Service) solutions, or even custom integration code can provide more streamlined and affordable alternatives for connecting modern systems.

Choose right tool for the right job 

This.