Top 5 Legal Issues in AI and Machine Learning

Brian Hall - November 15, 2017 - Artificial Intellience, Internet Law, Internet Lawyer

img

Bonus content: Can You Make Money Using ChatGPT? (See video below)

In modern times, life is less about keeping up with the Joneses and more about keeping up with the Jetsons. Technology has infiltrated almost every aspect of life, where a dead cell phone feels like a lost limb. With artificial intelligence (AI) and machine learning developments, battery-dead devices will begin to feel like lost companions. As an AI lawyer specializing in AI law, I am uniquely aware of how fast AI develops. ChatGPT by OpenA is the best current example of an AI service platform that is driving the explosion of public awareness about the potential of AI to change the way we live and work. Evangelists of the ChatGPT service believe AI will revolutionize customer service, legal services, healthcare, finance, retail, education, software programming, art, and more. 

Legal implications of artificial intelligence

Artificial Intelligence (AI) and machine learning refer to software that can adjust how their coding reacts to input over time as they “learn” more about the information they are receiving. Artificial intelligence affects life, from Siri to smart cars to online advertisements. The full range of rewards and risks that arise from using these technologies has not been fully explored. AI can automate any process, perform predictive analytics, improve information accuracy and do so in seconds. AI can also create better user experiences and provide more personalized services.

AI Law Is Undeveloped. AI Legal Issues Are Only Just Beginning.

There is no ‘category’ of law called “AI law” which we can apply to legal issues created and presented by artificial intelligence.  Lawyers, judges, and arbitrators will need to apply classic legal principles to AI. As with any new technology, the current caselaw and statutes may not easily be adapted to unique issues presented by artificial intelligence. It will take years for attorneys representing AI companies, judges presiding over cases involving AI, and arbitrators specializing in AI to figure it out. A lawyer advising an AI company or client will need to develop new law, as well as attempt to provide advice and recommendations about how the law will be applied to issues of copyright infringement, software as service agreements, drafting terms of use for AI applications and services, developing data privacy models and privacy agreements, and understanding AI copyright issues when the author of the work is a machine, as opposed to a human. 

Five AI Legal Issues Every AI Project Should Consider

At least five legal issues are innately associated with AI and machine learning.

  1. AI computes faster than Congress. Technology has been developing rapidly since the Industrial Revolution, quicker than the law can pace. So, when legal issues arise, more often than not, they are a case of first impression. Lawyers who have an AI case fall into their lap will be treading into uncharted territory without a map and trying cases in front of judges who may not comprehend the technology.
  2. Who is at fault? Trying to find the liable party is like playing a science-fiction version of Clue if an accident involves AI. If a smart car hits a pedestrian, who is the guilty party? The programmer in the office with the source code? The owner on the road with the car? The manufacturer in the lab with the testing protocols?
  3. When artificial outweighs intelligence. AI often has to identify objects such as cars or people. However, because AI relies on cameras and coding, things like contrast, color, and image density affect AI’s “thinking” much more dramatically than humans’. A person would not likely miss a white semi-trailer “against a brightly lit sky.” A human would not mistake a pattern of dots or lines for a starfish. AI also can reflect the developer’s biases, as seen in many software programs’ tendencies to develop racial biases.
  4. Humanizing robots. As technology develops, AI gets closer to actual consciousness. The United States has already granted rights and legal responsibilities to non-human entities, namely corporations; it is not unfathomable that robots and machines utilizing AI will be granted the same. Facebook has already created AI sophisticated enough to develop its own non-human language. Were these machines’ civil rights violated when Facebook shut them down? If AI commits a crime, can the software itself be held liable? Switzerland faced that very problem when a robot bought illicit substances online.
  5. Privacy no longer exists. AI already tracks and predicts individuals’ shopping preferences, political preferences, and locations. The data accumulated and shared between these technologies has already created many legal controversies. However, AI is starting to tackle more controversial subjects, such as predicting sexuality and the propensity to commit a crime. Will these predictions be able to be used in trials? Or will the AI serve as experts to be cross-examined to determine the validity of their opinions?

ChatGPT By OpenAI Takes Center Stage

You no doubt have heard of ChatGTP through the project OpenAI. ChatGTP provides a service whereby users can interact with a chatbot using natural language as though it is a normal human conversation. OpenAI is based on a deep learning model trained on millions of conversations, allowing it to learn and respond to questions and prompts (commands, context, approaches).  Even more critical, ChatGTP learns more with every conversation.  ChatGTP exploded into mainstream consciousness in late 2022 with the launch of version ChatGTP 3, turning the world upside down with excitement about the possibilities of artificial intelligence. Developers and others are creating services on top of the OpenAI codebase using the ChatGTP API. 

How Do I Get Permission to Use ChatGTP For Commercial Purposes (i.e., can I make money off ChatGTP?)?

Because ChatGTP is a service, there are no licenses to use the API; however, a ChatGTP service agreement defines appropriate use and the limitations on use. There is also an OpenAI Terms of Use agreement which you will need to review. Here is what ChapGTP says about the commercial use of its service by third parties:

(c) Restrictions. You may not (i) use the Services in a way that infringes, misappropriates or violates any person’s rights; (ii) reverse assemble, reverse compile, decompile, translate or otherwise attempt to discover the source code or underlying components of models, algorithms, and systems of the Services (except to the extent such restrictions are contrary to applicable law); (iii) use the Services to develop foundation models or other large scale models that compete with OpenAI; (iv) use any method to extract data from the Services, including web scraping, web harvesting, or web data extraction methods, other than as permitted through the API; (v) represent that output from the Services was human-generated when it is not; or (vii) buy, sell, or transfer API keys without our prior consent. You will comply with any rate limits and other requirements in our documentation. You may use Services only in geographies currently supported by OpenAI.

Note the terms of use do not exclude commercial use.  So, you can use ChatGPT to launch a business, monetize or make commercial use.

When it comes to AI and machine learning, there are currently more legal questions than answers. Artificial intelligence legal issues have yet to arise. But don’t worry; robots may have legal answers for us soon enough. When they do, will we be ready to listen? How does machine learning fit in Law? Law, including AI lawyers, is but one area to be disrupted by AI.


 

GET IN Touch

We’re here to field your questions and concerns. If you are a company able to pay a reasonable legal fee each month, please contact us today.