Insight
December 18, 2023
Software is essential to modern vehicles, especially electric and autonomous vehicles (EVs and AVs). Software controls a myriad of vehicle functions, including suspension, brakes, steering, power delivery, battery charging and discharging, user interface, media, internet connectivity, autonomous driving, speed and fuel monitoring, accident avoidance, lane positioning, and more. Almost a quarter of all modern automotive commercial applications are based on open source software (OSS), and this number is expected to grow to encompass many of these automotive functions. There are many open source automotive initiatives around the world that are leveraging the low cost and flexibility of OSS to work on a variety of projects for EVs and AVs. However, these benefits also come with significant risks. Automakers and other developers in the industry must keep this in mind as they continue to grow the suite of software that controls vehicles.
Introduction to OSS
OSS is computer software that is distributed with source code to make it available for use, modification, and distribution. Although OSS is “open,” it is considered an original work, subject to copyright, and can only be used in accordance with the license terms imposed by the OSS’s copyright owner. Violating this license can expose companies to claims of breach of contract and copyright infringement, leading to injunctions, monetary damages, and even contamination of a company’s own codebase with unlicensed third-party code. There is a possibility. Despite the risks, OSS exists essentially everywhere. In 2016, there were approximately 84 open source components for each commercial application. By 2020, that number had increased to 528.
Why use OSS?
OSS has many advantages. It was (and is) low-cost or free, rapidly deployable, modifiable and maintainable, and continually improved by the coding community. Manufacturers prefer using OSS because it reduces development costs, reduces production time, and accelerates innovation. However, as mentioned above, the risks of OSS can be significant. For example, vulnerabilities in OSS can compromise applications and expose sensitive information. Additionally, poor OSS license compliance can lead to litigation. Even more alarming, and of particular concern for EVs and AVs, poor OSS quality can impact vehicle and occupant safety.
To mitigate these risks, automakers and other developers are taking proactive steps, such as adopting policies and best practices to detect and address security vulnerabilities and avoid license conflicts. can be taken. Companies should also train developers on OSS risks and diligently track their use of OSS throughout the manufacturing pipeline.
OSS and IP litigation concerns
In open source litigation, there are three categories of plaintiffs: rights holders (those who own or have rights to copyrighted software covered by an open source license), non-rights holders (those who do not have rights but seek damages based on non-compliance with the terms) ) is. open source licenses), and other third parties who claim to have been harmed by the use of open source software in violation of statutory, regulatory, or other legal requirements. The main issues that these lawsuits typically raise are disputes, copyright preemption, terms and conditions, and damages (monetary or some form of injunctive relief or specific performance).
Rise of generative artificial intelligence and intellectual property risks
The past year has seen an explosion of advancements in generative artificial intelligence (AI) programs such as ChatGPT, but the automotive industry has seen an explosion in advances in generative artificial intelligence (AI) programs such as ChatGPT, which are rapidly expanding the automotive industry from vehicle navigation and driver assistance to design and manufacturing, fuel economy, and even predictive maintenance. . However, the use of this rapidly evolving technology comes with legal responsibilities. For example, when developing and training AI programs, companies need to be aware of the sources and limitations on the underlying datasets, which may have access restrictions or underlying third-party rights.
Automakers are also using generative AI to create new content such as product design, driver interfaces, maps and other navigational aids, and even marketing materials. Important legal issues are currently being discussed in court over this content, with litigation generally focusing on: Infringement claims in the use of datasets in AI training. Infringement claims for output produced. Ownership and enforcement of generated output. Copyright infringement claims typically include claims of both direct infringement (use in training or output substantially similar to the copyrighted work) and derivative (contributory and vicarious) infringement. there is. Apart from the motion to dismiss the secondary infringement claims, defendants in these infringement cases primarily argue that the use of protected content to train AI is “fair use” under copyright law. I’ve done it.
Both the Copyright Office and the courts have taken into account that whether AI output is protected by copyright at all depends on whether there is sufficient human copyright. For example, in a recent issue regarding copyright registration for a graphic novel developed in conjunction with an AI program, the Copyright Office granted copyright protection only for the human-authored portions, i.e., the entire text and compilation. The Copyright Office then issued a policy statement stating, “Going forward, the Copyright Office will decide whether an AI contribution is the result of ‘mechanical reproduction’ or an ‘original mental concept’ of the author.” We will consider it.” [the author] gave it a visible form. ”
As this case law evolves, automakers should remain cautious when using AI and consider implementing the following practices to reduce the risk of infringement.
- Train employees on the basics of copyright law.
- Before using an AI tool, make sure you fully understand the tool’s attribution and/or acknowledgment requirements.
- Be transparent internally about the use of AI tools and provide regular training on AI policies.
When using AI tools:
- Consider the sensitivity of inputs/prompts in real-time, as sensitive information can be shared, kept within the tool, or duplicated and identified in the AI output.
- Be aware that AI output can be compromised. Consider image/portrait rights and use available filters to reduce risk.
- Check the accuracy of the output.
- Consider the protectability of your AI output and make significant and noticeable changes to make it yours (note that it requires human authoring).
conclusion
The proliferation of OSS and the rise of generative AI in modern vehicles, especially EVs and AVs, means that the automotive industry is entering an era of change. As reliance on OSS increases, automakers must navigate a complex mix of benefits and pitfalls associated with this collaborative model. Similarly, as AI is incorporated into more aspects of the manufacturing pipeline, companies will need to have appropriate policies in place to avoid accidental breaches and ensure ownership and enforcement. Working with your attorney and taking proactive steps is key to mitigating these risks.