OpenAI’s Latest ChatGPT Upgrade Aims at Business Workflows

An artificial intelligence (AI) chatbot that remembers what you asked it to do last month? That’s the promise behind ChatGPT’s latest upgrade, which could push the technology beyond quick answers into managing long-term business tasks.

The update lets ChatGPT track vendor payments and purchase orders over time, but companies weighing this AI workflow assistant face a core challenge. These systems may be getting smarter, but they still mix up facts and misunderstand requests — leaving businesses to decide how much they can trust AI with their daily operations.

The update, dubbed Projects, provides a dedicated space to work on long-form tasks like documents or coding projects. It allows users to create, edit and iterate within a structured canvas alongside conversations in an attempt to improve focus and collaboration by keeping work organized and easily accessible during discussions.

“While AI will undoubtedly transform project management, it’s unlikely to completely replace dedicated software companies,” Hikari Senju, CEO of AI company Omneky, told PYMNTS. “Instead, we’ll see a convergence of AI and project management tools, where AI augments existing platforms with intelligent features, such as automated task assignment, progress tracking and risk prediction. This will empower project managers to focus on higher-level strategic decisions and creative problem-solving.”

Unlike standard chat threads, ChatGPT Projects provide a persistent canvas where work can be saved, updated and refined over time. This feature lets users seamlessly transition between conversations and ongoing tasks without losing context or progress. OpenAI was designed to support complex workflows, enabling better organization and smoother collaboration. The Projects feature aims to make ChatGPT more practical for professional and creative applications by keeping iterative work consolidated in one place.

“AI, like ChatGPT Projects, won’t replace project management software but will disrupt it, forcing these platforms to evolve,” Ryan MacDonald, chief technology officer at Liquid Web, told PYMNTS. “Businesses are driving demand for tools that not only track tasks but also offer context, suggestions and proactive problem-solving. The shift is from passive tracking to intelligent decision support, raising the bar for traditional tools. Companies unable to adapt risk losing relevance in this new era. AI will serve as the connective tissue, automating and optimizing processes across these platforms, ensuring they stay competitive in a rapidly changing landscape.”

ChatGPT’s new feature is part of a growing trend of AI tools aiming to help people tackle bigger, more complex tasks. These tools are designed to keep everything — documents, code, notes and conversations — organized in one place, making it easier to focus, collaborate and stay productive.

OpenAI isn’t alone in this effort. Microsoft Copilot now works directly in Word and Excel, helping users draft and edit documents or analyze data without switching apps. Notion has added AI features that assist with brainstorming, writing and managing projects in a single workspace. For coders, GitHub’s Copilot suggests and refines code in their development environment, saving time and reducing context switching.

But as these AI tools become more deeply embedded in workplace software, security experts warn of growing risks.

Suriel Arellano, a technology business consultant, told PYMNTS that there are well-founded concerns about whether competitors could access sensitive business data stored in ChatGPT Projects.

Arellano said firms should encrypt data that’s being sent to and retrieved from the AI tool, as well as data that’s “at rest.” They should also be based on a zero-trust security model. They should establish and manage access controls with an eye to the potential risks involved with not only human access but also third-party application access.

“It’s also important to remember that a “fortress” approach, where defenses are placed only at the edges of a system, isn’t sufficient,” he added. “And defenses placed solely at the level of access to an AI tool also need to be increased.”

Another hurdle facing companies using AI systems stems from the quick pace of operation. While AI can analyze spending patterns and make recommendations within seconds, this rapid-fire capability poses risks for businesses trying to maintain financial control. Many firms now grapple with balancing AI’s swift decision-making against needing a thorough human review of spending proposals.

“There are a few schools of thought in terms of what such a system might look like,” Arellano said. “One way is to have an extremely robust system for when the AI says “do this,” for anything that’s even close to the spending threshold, so that any time the AI suggests making a significant alteration in the way money is being spent, the system kicks in and a human being gives the order.”

The post OpenAI’s Latest ChatGPT Upgrade Aims at Business Workflows appeared first on PYMNTS.com.