If you have used ChatGPT and found it very useful, and are curious to see if there are other ways in which you can take advantage of AI in your organization, this post is for you. We will discuss enterprise use cases for Generative AI (GenAI), and cover key caveats you need to be aware of.
-
Alternative to informational search: One of the most straightforward use cases is to use tools like ChatGPT, Claude or Perplexity as an alternative to Google search. It can save you time having to sift through different websites, and give you an immediate concise response. ChatGPT like tools also have the benefit of being conversational - you can follow-up on your original query and response and engage with it as you might with say a support chatbot.
One caveat is that GenAI tools don't use visual tools very well. As an example, if you search for how to do VLOOKUP in Microsoft Excel, the first link in the Google search result is to Microsoft reference on VLOOKUP that shows detailed excel screenshots of how to go about using the function. Another caveat is hallucination - the tendency of AI to make things up that are not factually correct.
-
Software development: Of the many promising use cases for GenAI, software development has emerged as a key one. According to Anthropic's Economic Index research, 37.2% of Claude conversations are in the Computer and Mathematical category. There are many different ways of using AI for software development based on the type of query (automation - where you ask the AI to build something from scratch, vs augmentation - where you where you collaborate with it on development) and type of work (e.g. frontend vs backend). The majority, according to Anthropic based on Claude Code usage, are in automation and front-end development.
There are a few caveats worth noting for GenAI use in software development:
- Security: Tools such as Cursor make it very easy to generate hundreds of lines of code based on a simple text prompt, and without adequate review of the generated code, it is easy to introduce security loopholes, which can pose significant data and reputational risk if the generated code is used in critical production systems
- Intellectual Property: There could be legal concerns to address about code generated from these tools, with reference to the licensing limitations imposed by the source code used as data to train the AI tool.
-
Creative: This is a broad category that includes all types of creative content - design, marketing and advertising, media and so on. GenAI can serve as a great resource for generating ideas, or even reviewing content. Similar to software development, it can be used to generate new content from scratch, or as a collaborator during content development. Video generation tools such as Veo are getting really good at generating human-like content.
As with software development, one caveat is IP concerns for AI generated content. Another is misuse, such as with deepfakes and the risks they pose such as misinformation and conducting frauds and scams.
-
Office tasks: If you have to draft emails or review emails you have written, or summarize a meeting or long document, or structure a presentation or report, AI can help you with it, improving your efficiency by reducing time spent on these tasks. It is worth listing down all the tasks you do in a day and asking an AI tool which ones of them could be automated using AI
One caveat is the potential risk this poses of exposing confidential or proprietary information to the AI tool, such as when you ask it to summarize an uploaded document or a meeting where you discuss proprietary information
As an organization looking to use AI, it is important to understand the ways in which you are able to do so along with an understanding of the risks and caveats and how you can mitigate them. Done right, AI has the potential to bring in significant improvements to overall productivity. We are here to help you with choosing and safely implementing the right AI technologies for your needs.