Thomson Reuters AI Copyright Dispute Going to Trial
PLUS: Sample ChatGPT prompt for attorneys
Welcome, fellow attorneys!
Lawyers and machines are duking it out in court, while Congress tries to lay down the law on AI - get the legal tea on big players slap-fighting over copyrights and regulations in this week's newsletter! We've got lawsuits, legislation, ChatGPT cheat codes, and more legal drama than an episode of Suits, so grab your popcorn and enjoy the AI legal throw down.
On the docket today:
Thomson Reuters AI Copyright Dispute Must Go To Trial
U.S. Congress Cooks Up Sweeping AI Framework - But Will It Really Become Law?
Summary of the Blumenthal-Hawley Framework on U.S. AI Legislation
ChatGPT Prompt To Summarize Content
New here? → Subscribe
Image created using DALL-E.
Legal research bigwig Thomson Reuters is majorly cheesed off because upstart Ross Intelligence supposedly copied stuff from Thomson's primo Westlaw platform to build its own AI legal search biz. Thomson's all like, "Theft! How dare you steal our precious headnotes, you rogue!" But Ross is just trying to follow its tech bro dreams, responding "Chill, we just wanted to train our AI with your fire content, ya feel me?"
Thomson threw down the gauntlet and sued Ross way back in August 2020. The parties filed dueling motions for summary judgment but 3rd U.S. Circuit Court of Appeals Judge Stephanos Bibas was like, "Hang on now, let's not get ahead of ourselves" when he dropped his opinion this past Monday.
Even though Thomson's mad heated about this alleged stealing of its stuff, Bibas said there's still "genuine disputes" around whether Thomson's content can be copyrighted and whether Ross's use was fair. Dang pesky details!
Judge Bibas wants a jury to sort out all the messy facts at trial before he brings down the justice gavel. So Thomson's desire for swift vengeance has been denied, and Ross gets to keep fighting the power a little longer. Although, it looks like Ross had to hang up its shingle because of litigation costs. Thomson's probably hella salty about having to actually, like, prove its claims in court. But you know judges - they love their due process and weighing evidence.
For now this legal beef will keep simmering until the trial showdown (no date has been scheduled). Bummer for Thomson, but them's the breaks in the wacky world of AI and copyright!
Why it matters: This battle between legal heavyweights has major implications for AI development and copyright law. If Ross illegally copied, it could chill investment in using copyrighted content to train AI. But if its use was fair game, copyright law may need to bend to enable AI progress.
*I hope ya’ll get that the quotes included aren’t direct quotes.
Imaged created using DALL-E.
On Tuesday, September 12, the Senate Judiciary Committee bros held yet another hearing on oversight of AI and potential legislative approaches. At this point it's like their favorite new thing to talk about! Gotta show the voters they're on top of this hip new technology called AI.
This time the hearing was all about a proposed Bipartisan Framework for U.S. AI, introduced by Senators Blumenthal (D-CT) and Hawley (R-MO).
Their big idea is to create an Independent Oversight Body to license and regulate companies that use AI.
The Framework would also allow people to sue AI companies for pretty much anything - privacy violations, civil rights breaches, general harms...the sky's the limit! No way that could lead to a barrage of lawsuits that stifle innovation.
In summary, Blumenthal and Hawley are trying hard to show leadership on AI regulation. But their proposal kind of seems like political theater that's unlikely to actually become law. Still, nice effort guys! Maybe someday Congress will pass meaningful AI oversight that smartly balances risks and benefits. A voter can dream...
Image created using DALL-E.
According to Senator Blumethal’s (D-CT) website, the Bipartisan Framework for U.S. AI would:
“Establish a Licensing Regime Administered by an Independent Oversight Body. Companies developing sophisticated general purpose AI models (e.g.,GPT-4) or models used in high risk situations (e.g., facial recognition) should be required to register with an independent oversight body, which would have the authority to audit companies seeking licenses and cooperating with other enforcers such as state Attorneys General. The entity should also monitor and report on technological developments and economic impacts of AI.
Ensure Legal Accountability for Harms. Congress should require AI companies to be held liable through entity enforcement and private rights of action when their models and systems breach privacy, violate civil rights, or cause other harms such as non-consensual explicit deepfake imagery of real people, production of child sexual abuse material from generative AI, and election interference. Congress should clarify that Section 230 does not apply to AI and ensure enforcers and victims can take companies and perpetrators to court.
Defend National Security and International Competition. Congress should utilize export controls, sanctions, and other legal restrictions to limit the transfer of advanced AI models, hardware, and other equipment to China Russia, other adversary nations, and countries engaged in gross human rights violations.
Promote Transparency. Congress should promote responsibility, due diligence, and consumer redress by requiring transparency from companies. Developers should be required to disclose essential information about training data, limitations, accuracy, and safety of AI models to users and other companies. Users should also have a right to an affirmative notice when they are interacting with an AI model or system, and the new agency should establish a public database to report when significant adverse incidents occur or failures cause harms.
Protect Consumers and Kids. Consumers should have control over how their personal data is used in AI systems and strict limits should be imposed on generating AI involving kids. Companies deploying AI in high-risk or consequential situations should be required to implement safety brakes and give notice when AI is being used to make adverse decisions.”
(1) Dragos Tudorache, Member of European Parliament and Co-Rapporteur of the EI AI Act, met with several members of Congress last week to discuss AI regulations in the U.S.
(2) Ben Brooks, Head of Public Poilcy at Stability AI, shared on LinkedIn that “next week, the EU will enter a final sprint to decide the date of AI models under the [EU] AI Act.”
(3) Allen & Overy, a British multinational law firm headquartered in London, introduces AI chatbot to lawyers in search of efficiencies. I guess the legal industry is capable of embracing new tech!
ChatGPT Prompt To Help You Summarize Documents, Laws, Opinions
Warning: Use these prompts at your own risk! Always read and verify the accuracy of the outputs!
“Act as an expert in legal strategy and risk management. Summarize the provided content into an executive summary, including key bullet points outlining all issues, analysis, and conclusions throughout. Additionally, identify the legal issues present and analyze the potential risks and benefits of pursuing a legal strategy to succeed on these issues, considering costs and risk management. Please provide a friendly and jargon-free response in paragraph and bullet format.
Do you understand the instructions?”
Once ChatGPT confirms that it understands the instructions, it will prompt you to provide the content.
Follow up prompts:
Can you elaborate on the specific legal issues you identified in the content?
How can the costs of pursuing a legal strategy be managed effectively?
Are there any alternative approaches, such as negotiation or mediation, that could be explored before pursuing a legal strategy?
Meme Of The Week:
Photo credit: David Hookstead/Twitter
(if you don’t get this meme, you have to watch ChatGPT’s latest feature demo and know about the Rittenhouse trial)
How often do you use ChatGPT in your practice?
That’s all for today!
Catch ya’ll on LinkedIn: https://www.linkedin.com/in/veronicalopezmonterroso/
What'd you think of today's email?