Home > IP Express > Copyright

Microsoft’s Shield for AI Users Against Copyright Claims

Post Time:2023-12-01 Source:europa.eu Author: Views:

On 15 November, during Microsoft Ignite, its annual gathering for developers and IT professionals, Microsoft disclosed a significant expansion of its policy to protect enterprise customers from potential copyright infringement claims related to generative artificial intelligence (AI), specifically Azure OpenAI Service.

Under this new policy, customers who purchase a license for the service can rely on Microsoft to defend and indemnify them if they face a copyright infringement claim arising from their use of Azure OpenAI or the results generated by the service. Interestingly, this announcement closely mirrors OpenAI's previous announcement that it would cover the legal costs of customers facing copyright infringement claims arising from the use of its tools. For more details on OpenAI's announcement, we invite you to read our previous blog discussing, which you can see here.

Nevertheless, there are notable differences between the two policies. Importantly, Microsoft's enhanced policy does not automatically apply to all Azure customers. To take advantage of these new protections, subscribers must implement "technical measures" and comply with specific documentation requirements to mitigate the risk of creating infringing content using OpenAI models. Microsoft has not yet provided detailed information about these measures. In addition, users must use the "content guardrails and filters" built into Microsoft's AI offerings to maintain legal protection.

Recently, there has been considerable discussion about the relationship between Microsoft and OpenAI. In July 2019, Microsoft invested $1 billion (€916.5 million) in OpenAI, establishing Microsoft as the "exclusive" provider of cloud computing services to OpenAI. This partnership has been continuously expanded over the years with billions of dollars of investment from Microsoft.

A noteworthy distinction between the two companies is their approach to solving legal issues related to copyright. Microsoft, through Azure AI, is focused on developing technology specifically designed to detect and prevent the misuse of intellectual property through AI models. Azure AI Content Safety is a specific tool on this platform that reflects this focus on technological and active solutions to address legal issues related to copyright. OpenAI, on the other hand, is more focused on providing options and recourse for content creators affected by IP-related issues in generative models. Although OpenAI does not have a platform comparable to Azure AI, its activities are aligned with the research and development of advanced AI models. Its focus is on enabling content creators to control their training data and offer direct compensation. In short, while Microsoft, with its Azure AI, focuses on technological solutions for active detection and prevention, OpenAI takes a more creator-centric approach, offering options and direct compensation. Both are dealing with copyright issues, but with different strategies and approaches, in line with their respective priorities and areas of expertise.