When performing job duties, be careful about the inquiries you direct to artificial intelligence (AI)

znak towarowy

Is there artificial intelligence?

The term “artificial intelligence” refers to applications that perform complex tasks such as online communication. Artificial intelligence is implemented by language models such as GPT-3 developed by OpenAI. The GPT-3 language model is used in the currently very popular chatbot “ChatGPT”. This language model has been developed based on large datasets. ChatGPT can generate text and achieve a response style similar to that of a human. Due to the fact that a huge amount of data was used to “train” the above artificial intelligence and that the “machine learning process” is constantly being developed by OpenAI and ChatGPT users who direct millions of requests to this application supporting its continuous development or “learning process”, ChatGPT is becoming an increasingly popular tool in carrying out professional and official duties.

The accuracy of the answers obtained using ChatGPT and the speed with which the application provides search results encourages its users to use it more frequently as a tool for carrying out professional duties.

The usefulness of applications using artificial intelligence (AI) can be confirmed by the fact that this paragraph of the article was generated based on answers generated by an AI application.

Data entered by users of chatbots

When directing a query to an application that is a chatbot, the user must enter a query that is a command to perform a specific operation by the application. When entering a command, the user may enter data that will not be subject to legal protection, which will mostly apply to general commands whose results will take the form of general assumptions and statements, and which, if they are to be applied to the performance of a specific professional or official duty, may require some adjustment, modification, or supplementation, for which the user using the response obtained from the AI application will be responsible.

However, the AI user can direct the AI application to a command containing specific, narrowly defined data that may be subject to legal protection, such as trade secrets, professional secrets, or any other confidential information.

What legal rights do the creators of data entered into AI applications have?

The rules for using both commands entered by users of AI applications and responses generated by the AI application based on them are determined by the regulations and policies for using AI applications.

One of the services offering an AI application in its terms of use included provisions stating that both the data entered by the user into the application (input) and the results obtained by the user using the artificial intelligence application (output) can be used to improve the language model or the application to which they were introduced. This means that the creators of the application are entitled, in accordance with the application’s terms of use, to use the data entered into the application and the data generated by it in response to user queries to develop their artificial intelligence. The terms of use and policies of AI applications do not require their creators to keep confidential the information entered by the user into the application. Furthermore, since the information entered by the user into the application as well as the results obtained by the user can be used in the process of “further learning” by AI, it is possible that both the information entered by the user and that which was generated based on their query will be used by the artificial intelligence to provide an answer to another user who will ask a similar query in the future.

In the terms of use of another AI application, it is indicated that the user of the application grants its creators and cooperating entities with them the right to use both the information entered into the application and the obtained results, including – the user grants unlimited license rights, which include in particular the right to copy, distribute, transmit, publicly display, publicly use, reproduce, edit, translate and format content. The user also grants the right to grant further licenses by the owner of the application. According to the service’s terms of use, these rights are not subject to limitations. As in the case of the first chatbot application, the terms of use of the chatbot tool did not include an obligation for the creators of the application to keep the information provided by the user confidential.

The use of AI applications for business purposes may lead to a breach of legally protected information.

The above-described terms of use of AI services indicate that directing AI to perform, or analyze data that is confidential may result in the disclosure of this information, and thus lead to a violation of applicable laws or the provisions of an agreement requiring the preservation of certain information in confidence.

In addition, confidential information entered into the application can be used to develop responses for another application user, and in accordance with the terms of use of AI applications, users have no rights to the search results that other users will receive in response to their queries.

The above requires caution and great care when using these new and undoubtedly very attractive work tools. It may also be necessary to introduce appropriate mechanisms and procedures in the company/organization, whose aim will be the conscious and legal use of artificial intelligence applications by employees and collaborators using these tools to achieve business goals.