The European Union (EU) has recently taken measures to tighten regulations on the use of AI in content creation and has begun the process by creating a draft of a bill. After a vote taken on April 27th, the draft was approved, laying the groundwork for the classification of AI tools by risk level. Generative AI tools, such as ChatGPT and Midjourney, must disclose any use of copyrighted material inAI training according to the regulations.
Svenja Hahn, a deputy of the European Parliament, applauded the decision and claimed it will foster innovation and growth for the economy. This decision has coincided with the release of the European think tank Eurofi’s magazine, which includes a section solely dedicated to AI and machine learning applications in finance in the EU.
The EU’s Artificial Intelligence Act, the main bill driving these regulations, was proposed as draft rules almost two years ago. Regulation would be beneficial, according to Google Cloud director Georgina Bulkeley, when ensuring that AI technologies are handled safely.
Elon Musk, the CEO of Tesla and SpaceX, also recently threatened Microsoft with a lawsuit due to their use of AI trained on Twitter data. In response to this the EU’s data watchdog voiced concern over the potential issues US AI companies could face in light of GDPR.
ChatGPT is an open-source natural language processing (NLP) project supported by Microsoft, dedicated to creating an AI-powered conversational chatbot. MidJourney is a similar open-source NLP project backed by Square, which builds AI powered conversation bots as its main product. Together, these two AI tools make up the majority of the generative AI tools present in the EU technological landscape, and with the new regulations around the corner, they could be subjected to tougher transparency procedures.
The concerns regarding AI regulations and the usage of copyrighted materials have been rising due to the tech’s increasingly used for content creation. With the bill passed, owners and developers of AI technology will have to explicitly report on the AI’s use of copyrighted material and ensure the use is both legal and ethical. The caveat of this decision is that citizens can rightfully expect their online data to remain safe and confidential.
