Weekly Roundup: Artificial Legal Requirements
AI companies brazenly flout IP laws. Plus everyday incursions into personal privacy
All the news that’s fit to copy
The AI legal epic continues to unfold. Last week provided a few updates on ongoing cases as well as a couple of new lawsuits and regulatory investigations.
First, “Eight daily newspapers owned by Alden Global Capital sued OpenAI and Microsoft on Tuesday, accusing the tech companies of illegally using news articles to power their A.I. chatbots.” This lawsuit, while new, isn’t very different from other lawsuits accusing the AI giant of copyright infringement. It appears to be that these companies are taking calculated risks flouting IP law; A New York Times investigation last month found
OpenAI, Google and Meta ignored corporate policies, altered their own rules and discussed skirting copyright law as they sought online information to train their newest artificial intelligence systems.
Their situation is urgent. Tech companies could run through the high-quality data on the internet as soon as 2026, according to Epoch, a research institute. The companies are using the data faster than it is being produced.
“The only practical way for these tools to exist is if they can be trained on massive amounts of data without having to license that data,” Sy Damle, a lawyer who represents Andreessen Horowitz, a Silicon Valley venture capital firm, said of A.I. models last year in a public discussion about copyright law. “The data needed is so massive that even collective licensing really can’t work.”
The details of the copyright case will certainly hang on technical details of established IP law and a judge's ability to apply them to new technology that requires the pilfered material to survive. Whether or not the judges believe the arguments AI companies make in defense of their actions (or if the AI companies believe them themselves) is tangential to reality, at best. Much of the material on the internet has already been added to training data.
In a parallel effort to gain access to licensed material, OpenAI signed a deal with The Financial Times (FT) to use their material.
Under the terms of the deal, the FT will license its material to the ChatGPT maker to help develop generative AI technology that can create text, images and code indistinguishable from human creations. The agreement also allows ChatGPT to respond to questions with short summaries from FT articles, with links back to FT.com. This means that the chatbot’s 100mn users worldwide can access FT reporting through ChatGPT, while providing a route back to the original source material.
The benefits seem lopsided, given that OpenAI isn’t concerned with IP laws, and FT will gain access to 100mn readers. Part of the benefit to OpenAI may be the legitimacy FT will bring to ChatGPT answers; part of it may just be business strategy.
The deals highlight the uncertain balance of power between AI and the media. On the one hand, uncertain copyright protections and the easy access to material online has encouraged many AI companies to take the chance with unlicensed data, hoping they will be able to claim fair use in any legal battles. When they do need to license material, the commodity nature of much reporting encourages a “divide and conquer” approach – if only one deal is needed to keep a chatbot up-to-date with the latest news, this offers strong bargaining potential.
Google-opoly
A judge heard the closing arguments for an anti-trust lawsuit against Google: "The government argues Google illegally cemented a monopoly in search by paying Apple and other tech partners billions of dollars to feature the Google search engine in their products.”
The timeline for this trial and other anti-trust tech trials has been glacial.
By the time closing arguments [for the Google trial] commence this Thursday, May 2, nearly 6 months will have passed since the conclusion of those headline-stealing proceedings.
The Federal Trade Commission is pushing for a 2024 trial date in a case filed in December 2020 against Meta’s alleged social media monopoly.
The FTC’s case against Amazon is expected to go to trial in 2026, three years after the filing of the FTC’s long-awaited case against the multinational conglomerate.
It may not be apparent to most how quickly trials have moved in the past, given how long the appeals process can drag out; however, juxtaposing the current trials against the 1998 anti-trust trial against Microsoft, it is shocking:
The case against Microsoft was filed on May 18, 1998. On May 22, a trial date was set, and trial commenced 5 months later on October 18, 1998. If you put the timeline of the Microsoft case against that of the current case against Google Search, the Microsoft case had already been appealed to the DC Circuit, remanded to the lower court, and almost completely resolved before trial would have begun in the Google Search case.
Given that this is the first of a few FTC anti-trust suits, it will undoubtedly provide momentum, if not precedent, to the ones still making their way to the docket.
Not-so-personal vehicles
Last month, I began a new installment of my “Inside the Covered Wagon” series, shifting from the previous topic of Autonomous Weapons to Privacy and Anonymity. The first part was a privacy primer; the next part will be a piece on how consumers outsource their privacy. A perfect example of this phenomenon is captured in several articles last week on vehicular privacy. According to the NY Times, some of the more blatant violations:
In a March report, the [Government Accountability Office] concluded that consumers don’t fully understand crash avoidance technologies and driver support systems, the improper use of which “can compromise their safety benefits and even pose a risk on the road.”
The Federal Communications Commission and California lawmakers want to prevent mobile car apps from being used for stalking and harassment.
An investigator from the [FTC] reached out to a man named in a New York Times article whose insurance premium increased after General Motors provided data about his driving behavior to the insurance industry.
Automakers generally retain a car’s location information for years — as long as 15 years in the case of Hyundai.
The privacy shortfalls with cars and other ubiquitous tracking devices have been used by stalkers and domestic abusers.
There was other emerging tech-privacy news this past week as well.
“Microsoft has reaffirmed its ban on U.S. police departments from using generative AI for facial recognition through Azure OpenAI Service.”
Section 702, the controversial section of the Foreign Intelligence Surveillance Act, swept up nearly triple the number of U.S. identities than last year. “Over 30,000 U.S. persons, companies or other entities were unveiled in surveillance reports viewed by or shared among law enforcement officials and other agencies, up from some 11,500 in 2022, according to the annual report from the Office of the Director of National Intelligence released Tuesday.” This increase is partly due to increased cyber attacks against critical infrastructure.
“The [FCC] on Monday said it fined the largest U.S. wireless carriers $200 million on grounds that the companies sold their customers’ location data to third parties without consent while not taking steps to protect that info from security compromises.”
In other news
The Biden Administration announced that all 180-day action items from their AI Executive Order were completed on schedule.
NIST launched NIST GenAI, a program that “will release benchmarks, help create “content authenticity” detection (i.e. deepfake-checking) systems and encourage the development of software to spot the source of fake or misleading AI-generated information.”
“Ukraine on Wednesday presented an AI-generated spokesperson called Victoria who will make official statements on behalf of its foreign ministry.” The spokesperson was created to save time and resources.
An extremely popular and effective benefit that subsidized internet for low-income individuals faces death in the House because the renewal bill can’t get out of committee.
Before you go
Wired published a great interview and profile of Arati Prabhakar, the director of the White House Office of Science and Technology Policy and the woman advising President Biden on all things emerging tech.
