Microsoft and ChatGPT: stay ahead of the integration, stay on top of the game

by Javier López-Guzmán - Compliance Consultant
| minute read

The superfast development of generative artificial intelligence (AI) tools, such as ChatGPT, has shaken the grounds of corporate thinking in recent months. Because of their undeniable impact at an operational and managerial level as well as at the much more publicised ethical level, these tools are clearly a gamechanger for business. The adaptation capabilities for their use will mark the corporate development of any enterprise or institution, in almost all sectors of our economy.


Taking this inescapable new trend into consideration, Sopra Steria proposes one quick and direct way of adapting to the use of generative AI tools: the integration of ChatGPT and Microsoft software. For one thing, Microsoft’s global leadership in today’s desktop applications market is undeniable. And its software is present in the everyday use of all these applications.


Which is not to say that Google Workspace and Apple iWork users won’t be confronted with this new AI development soon enough. Perhaps even an open-source alternative for generative AI tools, such as Apache OpenOffice used to present for desktop applications, will also arrive eventually.


Privacy concerns

The importance of adapting your business to the arrival of these generative AI tools cannot be overstated. Microsoft has already set the first step in this adaptation game by integrating ChatGPT into its Bing search engine. And it has to be said: the (currently) most powerful generative AI tool fits rather well with Microsoft’s global-reaching search engine tool.


One precaution with regard to data confidentiality should nevertheless be noted as, on the one hand, the way these AI tools are developed and operate makes them extremely hungry for data. Whereas, on the other hand, the protection of personal data and the confidentiality of information are not yet guaranteed, unfortunately.


This deficiency has already led certain national regulators, such as the Italian Data Protection Authority and the Spanish Data Protection Agency, to flag privacy concerns over the use of generative AI. On the European level also, a specific taskforce within the European Data Protection Board (EDPB) is set to be established. Even prominent figures from the world's leading tech communities in Silicon Valley have raised some ethical concerns in an open letter to the public.


Human-like interaction

Of course, the market is not limited exclusively to the Microsoft search engine Bing. A fight for market dominance is bound to occur, involving alternative solutions by a number of competitors, both old and new. One such competitor is Google with its AI chatbot service Bard.


As a result of that industry battle, the world of information searches on the Internet is set to change fundamentally. The use of natural language AI tools, for example, will supplant today’s traditional and common ways of interacting with internet search engines, such as by clicking a mouse, tapping on a screen, or issuing simple one-line voice instructions. Instead, we should be able in future to conduct our searches and state our specific needs through a more organic, human-like interaction.


In terms of the protection of our personal data, this fundamental change in the way we use or interact with search tools will elevate even further the amount of data collected from these tools. Search engines per se will disappear to leave the path for virtual assistants that follow us online in every interaction.


In terms of business, the reliability of the information sources used by these new search tools will be the key challenge to gain actual benefits or business value from them. Chat GPT, for instance, offers a single response only to information requests: the one response it considers the most relevant to answer your prompt.


Prompt engineering

Microsoft’s integration of AI tools does not finish here, though. The objective of the company is to fully integrate not only ChatGPT but also all other generative AI tools in its software. An example is the already existing OpenAI GitHub Copilot, which turns natural language prompts into coding suggestions across dozens of language. This makes it an ideal development tool for regular office [JDP1] users who cannot code. For the protection of personal data, however, identifying the data flows and origins in these tools may be more difficult than with previous digital methods of processing.


The most important success factors for information gathering, now more than ever, are the right choice of sources and the guarantee of trustable research. Awareness about the training of the AI tools and algorithms used behind each search engine is very important, in order to be able to trust its sources.


The integration of generative AI tools will develop soon now, in a tight timeline. When it is completed, Microsoft tools like Word or Excel used by white collars every day will have changed fundamentally. And prompt engineering will be a new skill to show off in business relations.


Looking for expertise and counsel to guide your implementation of AI tools in the best compliant way? Sopra Steria has the skilled technical experts ready for it, from digital architects to cybersecurity engineers. Our portfolio includes information management capabilities, business continuity, analysis of personal data impact, and privacy by design embedded in architecture. Contact me or my colleagues for a no-obligation consultation.