Looking into the AI crystal ball
14 March 2023
by Linda Sheehan
With 2023 in full swing, it is clear that AI will continue to play an increasingly significant role in the business landscape. The demand for AI solutions and data-driven insights is on the rise, as businesses of all kinds seek to make more informed decisions and gain a competitive edge.
One emerging trend is the integration of traditional legal services with “legal and compliance as a service,” with a particular focus on eDiscovery and contract tech services. In this context, ChatGPT, a generative AI model, has garnered much attention for its ability to generate human-like responses to complex questions and tasks.
As we explore the potential of AI in various areas, such as product development, information searching, law, and business operations, we must also weigh the benefits of increased efficiency against the need for creativity, innovation, and the human touch.
ChatGPT, has facilitated an explosion in interest from individuals that had not previously considered incorporating AI into their day-to-day practice. Bill Gates, cofounder of Microsoft, says ChatGPT is as big an invention as the internet. Microsoft recently made its third investment in ChatGPT of USD10-billion.
In light of this, our 2023 predictions are based on AI interactions.
Many AI-based products have failed due to the lack of front-end useability. In other words, the interaction between humans and machines has been too complicated and the machine needed too much technical direction for common-place use.
ChatGPT’s ability to understand questions from humans, analyse vast amounts of data to generate accurate and human-like responses is perhaps the missing link between products on the market now and the future. At the very least, it has made advanced technology seem friendlier and more approachable to business.
Low code, no code platforms allow subject matter experts to create powerful digital applications without deep IT technical expertise. API integrations, which can enable a business’ current disparate tech stack to talk to each other, combined with cloud security and scalability means that we expect faster and better evolution of innovative solutions to streamline and improve operations.
In the short time since ChatGPT3’s November 2022 release, the way we search for information has already changed. There is no need to get left behind in a meeting because you don’t have the same background on a topic as everyone else.
The release of ChatGPT has been seen as a potential threat to Google’s search engine because it provides a more user-friendly way to search for information. ChatGPT provides quick answers in simplified terms drawing on millions of data points that it has been trained on. It has proven to be a worthy research assistant and content creator. It has also been used for troubleshooting for developers. This is in contrast to Google providing a list of websites to read through but not, for example, the actual code a developer needs.
In response, Google CEO Sundar Pichai declared a Code Red and released Bard. Bard AI service is described as an experimental conversational AI service. On its release, ChatGPT had been trained using the internet up until 2021. Bard draws information from the web making it more up-to-date and thus relevant.
Google Bards’ release highlights the importance of not rushing in too soon. Bard’s first public demo produced a factual error in early February 2023 which reportedly resulted in Google’s parent company, Alphabet, losing USD100-billion in market value.
ChatGPT is now available in Azure Open AI Service. The release of the GPT-4 language model is imminent according to an announcement by Microsoft Germany CTO Andreas Braun at an AI kick-off on 9 March 2023. GPT-4 is expected to move from the written word responses of its predecessor to responses through audio, images and video. It is anticipated to create AI-generated videos from simple text prompts, expanding its use cases to movie production and eLearning training. Some report that GPT-4 will feature facial recognition to study and read human emotions.
We expect that generative AI, in time, will improve the results of the current tried and tested tools. The American Bar Association has previously urged (ABA Resolution 112) that, “….courts and lawyers to address the emerging ethical and legal issues related to the usage of artificial intelligence (‘AI’) in the practice of law”.
- Judges and lawyers are already using generative AI to assist in drafting briefs and a Colombian judge has told media that he used ChatGPT to search for precedents and also to draft his ruling more quickly. As long as other research tools were used to ensure up-to-date and accurate information. This is a great example of using technology for efficiencies with a layer of subject matter expertise and fact checking to ensure the accuracy of the outcome. Colombia has also hosted its first-ever court hearing in the metaverse.
- Machine learning is already judicially approved for finding relevant documents for discovery purposes in litigation and widely used in extracting contract clauses for due diligences.
We anticipate a rise in data privacy and security concerns. Employees are feeding ChatGPT with client sensitive information to perform a task for it without knowing where this data is being transferred and stored resulting in data privacy and IP risks. Similarly, if a corporate’s outside counsel is using ChatGPT, it should consider any unintended waiver of legal privilege.
According to Dark Reading, “In one case, an executive cut and pasted the firm’s 2023 strategy document into ChatGPT and asked it to create a PowerPoint deck. In another case, a doctor input his patient’s name and their medical condition and asked ChatGPT to craft a letter to the patient’s insurance company.” Unfortunately, these are not rare instances, in the same article they cite, “Cyberhaven detected and blocked requests to input data into ChatGPT from 4.2% of the 1.6 million workers at its client companies because of the risk of leaking confidential information, client data, source code, or regulated information to the LLM.” We do not anticipate that robot lawyers will be representing us in courts any time soon. The overwhelming response from the legal fraternity is that it has its place in litigation and trial work, but it is not ready to have it representing clients.
A company, DoNotPay, made an offer on social media to pay USD1-million to any lawyer willing to use its AI “robot lawyer” in an upcoming case in the US Supreme Court at the beginning of 2023. Its robot lawyer was due to represent a defendant on a speeding fine in a New York court for the first time in February 2023, but promptly dropped the case and realigned its focus after purportedly receiving threats from the State Bar prosecutors. Concerns were raised about whether courts would allow the use of an ear piece that the self-represented party or lawyer would wear to communicate with the robot lawyer and thatit is a felony to practice law without a licence.
We anticipate a rise in lawsuits due to advancing AI, for example, in replicating images and text. Microsoft, GitHub and OpenAI is already facing a class action accused of violating copyright law for failing to provide credit for licensed code being used in responses. In March 2023, a class action complaint was filed in the US against DoNotPay in relation to claims that it has engaged in unauthorised practice of law by holding its “robot lawyer” out to be an attorney and selling legal services when it is not licensed to practice law.
ChatGPT provides the following examples for its own use cases in a business:
- automated reporting;
- curating knowledge bases;
- lead generation and market analysis;
- research and development;
- incident management reporting;
- ensuring compliance with processes and policies with relevant laws and regulations;
- data analysis to help make data-driven decisions; and
- process automation for financial reporting.
ChatGPT is still in the early phases of learning. It can make up false information and is based on old data. There are many tools and techniques available to assist with this list of use cases that may not be as advanced in generative AI but have been tried and tested.
Business should consider what problem it is trying to solve and how to generate return on investment. It must ensure that a full risk assessment is performed, including its security posture and policies and procedures. The appropriate approvals internally and externally must be obtained before its commercial application.
Business gains in productivity and reduced human prone errors must be balanced with retaining creativity, innovation and client centric services. Too much reliance on AI could be detrimental to progression and result in generic services. These tools should be used to maximise efficiency in manual tasks but not as a final answer to strategic business problems and client interaction. The human touch must not be lost.
Head of intelligENS
We are a team of focused legal and digital data analysis experts who have the know-how and experience to, firstly, import and index relevant information, and secondly, inspect and interrogate the information to deliver the necessary insights to you and your business, in the shortest possible time.
ENSafrica’s intelligENS team implements in-house AI machine learning technology ensuring precision and reducing costs.
ENSafrica is the first law firm in Africa to adopt RelativityOne in February 2022. RelativityOne is a globally leading document review platform with various document management and AI enhancers.
We have the ability to integrate and build upon the platform meaning it covers a broad spectrum of legal and compliance needs than eDiscovery. Read more about how we have used RelativityOne to translate urgent documents in their native format within 24 hours, saving time and cost for its clients.