Over the past decade, the field of AI has grown spectacularly, with many ground-breaking milestones being reported, such as a reinforcement learning agent that learnt to play ATARI games, AlphaGo that beat the best Go player in the world, reinforcement learning agents that learn to mitigate epidemics, deep learning models that predict protein structures and large language models that demonstrate interesting capabilities. While research in AI has traditionally been driven by academia, in the recent years, research companies, such as Deepmind and OpenAI, have been involved in significant breakthroughs. While in academia, there is a culture of keeping the reproducibility of scientific methods in high regard, which often goes hand in hand with making algorithms available as free or open-source software, this is not always the case in research companies. In this talk, we explore the free and open-source software ecosystem in the AI research community and contrast this with closed source applications, of which we discuss the ethical implications from the perspective of the scientific community. As research companies such as Deepmind and OpenAI explicitly state that their mission is to develop Artificial General Intelligence (AGI), we look ahead whether the first AGI instance will be free or open-source software and discuss what the implications might be for the scientific community and society in general.