OpenAI launched ChatGPT Enterprise as a more secure version of its ChatGPT large language model. Specifically, ChatGPT Enterprise is designed to be more secure and compliant with corporate privacy regulations than the prior versions of ChatGPT/GPT 4.0, supporting features like SSO and usage insights.
The enterprise AI market is divided on OpenAI, with some companies eager to adopt the technology and others hesitant. Enterprise sentiment on using OpenAI and its GPT-suite of products can be categorized as follows:
These companies are eager to leverage OpenAI’s offerings and are already doing so. Clearly, this group is inspired by this year’s remarkable focus on AI, and are willing to get on board and “learn as we go.”
Also see: ChatGPT Enterprise: AI for Business
Other companies are waiting to see how the market for generative AI offerings develops, and will choose their model accordingly.
They may have some concerns about data privacy or are not yet convinced how the technology will impact their current business processes. A “wait and see” approach is further bolstered by the budget cuts many companies have put in place over the last 12 months.
The remaining companies have already decided they will never use OpenAI. This is typically due to data privacy, compliance or business protection reasons. However, these “nevers” may arguably shift their stance as some of these data concerns are addressed, and/or the urgent need to deploy AI drives decision makers to take some calculated risks.
Specialized use cases
This includes many industries with highly specialized focus, including healthcare, construction and legal. It is unclear when or even if OpenAI will start focusing on these specific areas. Some companies would prefer to start with an open source model and focus on training that for specialized purposes.
While ChatGPT Enterprise does guarantee they are not using data from customer prompts to train OpenAI models, there have been other solutions that offer similar guarantees, including from OpenAI’s partner Microsoft.
For some firms, licensing OpenAI could trigger a new process which would require all existing clients to sign off on new terms. The cost of doing so might outweigh the incremental product improvement gained from adopting OpenAI.
Betting on faster development and the future of open source
Open source models are rapidly advancing, and OpenAI still does not allow access to certain methods like Reinforcement Learning from Human Feedback (RLHF).
While open source models are currently behind GPT 3.5/4.0, they are improving every month and some industry experts believe they will catch up to commercial offerings in 2024. Given the timelines for enterprise clients, they would rather place a long-term investment in developing their own model rather than adopting OpenAI for the short-term.
Also see: Top Generative AI Apps and Tools
Worried about OpenAI’s legal troubles
Corporations such as The New York Times and Barry Diller’s IAC, and individuals such as authors and stand-up comedians like Sarah Silverman, are suing OpenAI over scraping their data to train OpenAI.
While the outcome from these lawsuits are far from a given, some companies are hesitant to bet their future business and product offerings on an API whose legality is being called into question.
Doubts about changing performance issues
While OpenAI is making rapid improvements to its models like ChatGPT Enterprise, many are also observing that its capabilities have changed over the last few months. Even beyond the issues of hallucinations and non-deterministic answers, it’s challenging to build a product on a foundation that may change for reasons outside your control or knowledge.
Bottom Line: The Future of ChatGPT Enterprise
Overall, the enterprises remain split on integrating OpenAI in their business model. The launch of ChatGPT Enterprise may convince some companies to switch to OpenAI, but there are still a number of factors that could prevent adoption.
In the face of these challenges, OpenAI remains a leading player in the enterprise AI space. In a recent survey, of 1,000 general population adults who have heard of artificial intelligence, ChatGPT was recognized by 49% of respondents (although only 22% said they had heard of OpenAI).
The company is continuing to improve its models and expand its offerings, and it is likely that OpenAI and generative AI as a whole will continue to gain traction across industries in the years to come.
Read next: Generative AI Companies: Top 12 Leaders
About the author:
Ivan Lee is the CEO of Datasaur