The AI & Legal Update #15 - ⚡️ Illegal AI Activities & Happy Babies ↵

Hi, everyone. It’s January 2025, and we've got AI policy announcements, state AGs ready to party, crackdown on the misuse of AI, a $615/hour expert submitting fake cases (gah), and a bonus clapping baby video.

This week, your editor spoke at the Bar Association of San Francisco’s In-House Counsel Conference. Three interesting quotes ...

On Loper Bright - "Companies! It's a great time to challenge *anything*. What are you challenging?"

On CDA Section 230 - "Courts are breaking down whether Section 230 applies at the individual feature level in tech products. Good product counsel do this too."

On cybersecurity, by a former CISO - "This isn't PC but ... at the beginning of almost every security breach is a stupid person. Who clicked something."

📘 Read time: 6 minutes 17 seconds

 AI Regulatory

⚖️ Oregon state Attorney General Ellen Rosenblum (pictured below) issued guidance on how existing Orgeon law applies to AI. The report covers consumer protection, privacy rights, and anti-discrimination measures while acknowledging more legislation will likely come in 2025.

A few weeks later, Attorney General Rob Bonta (top photo) of California issued similar guidance that existing California law like the UCL and Unruh Act applies to AI.

🤼 So which AG wore it best? GC AI says “California's guidance is more comprehensive and aggressive, covering a wider range of topics and providing detailed explanations of how AI intersects with various areas of law.”

🏥 California also issued guidance on AI & Healthcare, explaining that “Healthcare-related entities that develop, sell, or use AI systems must ensure that their systems comply with laws protecting consumers. This requires understanding how AI systems are trained, what information the systems consider, and how the systems generate output.”

🔒 Texas AG Ken Paxton launched investigations into Character.AI, Reddit, Instagram, and Discord over children's privacy practices under state laws, following significant privacy enforcement actions against TikTok, GM, and Meta.

🔐 Related, the FTC shared guidance on the risk of consumer harm with the increase in AI usage. The tips they share include ways to make sure an AI product has safeguards for consumer protections and privacy. Must read for product counsel - see GC AI takeaways here.

Big Tech

🤑 OpenAI’s shift from non-profit to for-profit has started as their Board sees how capital intensive it will be to scale the company. 

💭 Speaking of, OpenAI CEO Sam Altman shared reflections from ChatGPT's launch to his unexpected firing and return, while expressing confidence in achieving artificial general intelligence by 2025 and emphasizing the responsibility of developing superintelligent systems. People are noticing Altman said that artificial general intelligence (AGI) would be achieved in the next couple years. Altman also reminisces ...

“We always knew, abstractly, that at some point we would hit a tipping point and the AI revolution would get kicked off. But we didn’t know what the moment would be. To our surprise, it turned out to be [the launch of ChatGPT in November 2022].”

[Editors note: your editor was opposite OpenAI on a deal in late 2021, and had the jaw dropping moment of knowing AI was finally here in a Feb 2022 demo, when an earlier version of GPT wrote an app to track calories given only a 6 word instruction. I was so taken at the time it motivated me to leave my GC role and found GC AI!]  

⛔️ Product Counsel beware? Plaintiffs accuse Meta's in-house lawyers of illegal activity … advising engineering teams on AI training? Through Cooley, Meta defends that the crime-fraud exception shouldn’t apply because fair use is not illegal (!) and that whether AI training qualifies as fair use is a novel question. See link to pleadings and GC AI explanation of the case and the CF exception here. Another fun fact? Zuck’s December 17th deposition in the case was 4+ hours on record and in Kauai (Kau-AI?). 🌺

💰 Elon Musk's xAI secured $6B in funding. Other AI companies are raising too, see infographic below.

Intellectual Property

📕 The House released the "Bipartisan Task Force Report on AI” with 66 key findings and 89 recommendations, organized into 15 chapters. The House is careful not to assign itself too big a task, and notes: “it is unreasonable to expect Congress to enact legislation this year that could serve as its last word on AI policy.”

💻 Check out this nice list of the top 6 legal considerations for licensing AI software. Issues to consider include different types of AI model licensing (public vs. private), data processing vs. training rights, and the nuances of model ownership and copyright issues.

🎶🎹 Anthropic agreed to implement AI guardrails while leaving a training data dispute unresolved in music publishers' lawsuit. The agreement includes protective mechanisms like content filtering and watermarking, but keeps specific technical details under wraps as the case continues in the N.D. Cal. See this GC AI chat for analysis.

AI & Legal

📘 Deloitte talked with 400+ CLOs for the 2024 CLO Strategy Survey Report. 93% believe generative AI can bring value to their organizations in the next 12 months.

🧑‍🏫 Judges are grappling with first-of-their-kind cases relating to privacy and IP in AI. Dechert's Brenda Sharton discusses educating judges on AI and points to the importance of educating the public and keeping the perception positive.

🔮 65 legal experts shared their predictions for Legal Tech and Legal AI in 2025, including your editor who predicts that “prompts are the new templates.” 

🤖 The Florida Bar President Roland Sanchez-Medina Jr. put out guidance for Florida lawyers on AI, emphasizing that lawyers “must uphold our duty of competence, regardless of the tools we use: and that “AI may be able to draft briefs, conduct research, and assist in legal analysis, but the lawyer remains responsible for the final product.”

🫠 A Stanford "expert" on AI submitted fake AI-generated cases in, ironically, a case involving an AI misinformation law. See pleading. "Professor Hancock believes that the AI-hallucinated citations likely occurred when he was using ChatGPT-4o to assist with the drafting of the declaration", and he explains that GPT-4o replaced his notes to "NTD: insert citation" with fake citations. People? Can we PLEASE (a) use a legal-dedicated AI like GC AI or, heaven forbid, another legal AI, and (b) take an AI prompting class so we know not to do that? PS - if you have taken one of the GC AI legal AI prompting classes, you are officially a better AI prompter than a Stanford phD in the field.

AI Globally

🛂 The Biden Administration announced a new AI export control framework - the “Interim Final Rule on Artificial Intelligence Diffusion” to “refine its framework for applying export controls to regulate the global diffusion of the most advanced artificial intelligence (AI) models and large clusters of advanced computing integrated circuits (ICs) to protect U.S. national security and foreign policy interest.” Seventeen EU nations face export restrictions under the rule. 

🖥️ AI chipmaker Nvidia came out swinging that they don’t like it, saying the guidance is “misguided” and that “[w]hile cloaked in the guise of an ‘anti-China’ measure, these rules would do nothing to enhance U.S. security.” GC AI scores Nvidia’s blog as spicy. PS - a few readers mentioned not liking the political talk, but it’s hard to report on AI without noting just how much big tech has re-professed love for the Trump Administration including specifically on AI:

• “The first Trump Administration laid the foundation for America’s current strength and success in AI.” - Nvidia

• “ The country has a unique opportunity to pursue this vision and build on the foundational ideas set for AI policy during President Trump’s first term.“ - Microsoft 

• “I'm optimistic [ ] with president Trump  [-] he just wants America to win”, Zuckerburg, who is also hosting a reception for the inauguration. See Axios’s nice breakdown of how Meta courted Trump.

💻 Another AI chipmaker, Taiwan-based TSMC, has reported its profits in the last quarter increased by 57%. It remains to be seen how new guidance from the U.S. this week that would further restrict exports of chips used for AI will impact foreign chipmakers. The new rules cap the number of AI chips that can be exported, however gives broader access for about 20 close allies including Taiwan.

AI & Corporate Adoption

⏰ Linked says AI will make shorter workweeks? Hmm. See LinkedIn’s “25 Big Ideas that will change our world in 2025,” including that businesses will need to focus on identifying AI applications that can boost their bottom line or risk losing their competitive edge. Plus, business leaders will shift from AI hype to focus on practical applications with measurable outcomes. Like, ahem, GC AI? 

🏢 OpenAI is hiring employment law and workplace investigations attorneys. This could be an indicator of the growing complexity with employment data security in an AI world. 

🏧 The banking industry could lose as many as 200,000 jobs globally from AI, Bloomberg Intelligence suggests. That's based on a survey of 93 CIOs and CTOs at major banks, including Citigroup and Goldman Sachs.

🤑 OpenAI put out an economic blueprint for AI. 

New, Notable, Random & Fun

💵 This National Law Review article shares key takeaways from The US Department of the Treasury’s report on Artificial Intelligence in Financial Services.

🛍️ According to Salesforce, $229B of global online sales (19% of total sales) were influenced by AI and AI agents, with shoppers using AI product recommendations, targeted offers, and customer service support bots. Editor’s note: AI told me about Walmart’s amazing Birkin dupe. Real thing one day! 

👏 🎥 Here's an extra dose of fun. Your editor took Runway's AI video tool for a test drive with a picture of my daughter and the prompt "Kira claps her hands happily". In seconds, the video animation on the right was generated (see the screen grab below). Pretty impressive! But ... can you see how her hands are a little adult-like and stop touching at the end?

PS - Runway product counsel, nice job on the disclaimer too (below). Notice the alternative references to the consumer terms or enterprise agreement.

Upcoming Classes

👩🏻‍🏫Advanced AI Prompting for In-House Legal - 201 on Thursday, January 23rd at 9 am PT/12 pm ET. Sign up here.

👩🏻‍🏫AI Prompting for In-House Legal - 101 on Tuesday, January 28th at 9 am PT/12 pm ET. Sign up here.

How did you like this newsletter? What can we do better? Click to send feedback. 

❤️  I loved it!

😐 Good but could be better. 

👎 It was meh.