Greetings and welcome to TechScape! After this newsletter goes live, you might find yourself captivated by the wedding snapshots of Jeff Bezos and Lauren Sanchez, the most glamorous pairing in the tech news sphere this year. I found the event to be both sticky and monumental. Although everyone attended, Charlize Theron wasn’t on the guest list; as I mentioned earlier: “We might be the only ones not invited to Bezos’ wedding, but that’s okay.”

AI Companies Begin to Prevail in Copyright Disputes

Recently, the tech sector achieved multiple victories regarding the usage of copyrighted materials for developing artificial intelligence products.

A noteworthy judgment from a U.S. judge concluded that the creator of human-like chatbots, which trained on books without securing author consent, did not breach copyright regulations. Judge William Allsup equated the act of using human writings to “readers aiming to become writers.”

The following day, a ruling favoring Meta emerged: U.S. District Judge Vince Chhabria in San Francisco concluded that the plaintiff did not provide adequate proof that AI technology from these firms would lead to “market dilution” by inundating it with similar works.

On that same day, while Meta gained a favorable ruling, a group of authors sued Microsoft, accusing the company of copyright infringement linked to its Megatron text generator. Given the rulings favoring Meta and the chatbot sector, authors are facing a challenging uphill battle.

These cases are minor skirmishes in a larger legal struggle surrounding copyrighted media. Just three weeks ago, Disney and NBCUniversal filed a lawsuit against Midi Joanie, claiming its AI image generator and upcoming video tools unlawfully utilized iconic characters like Darth Vader and the Simpsons. Meanwhile, major record labels—Sony, Universal, and Warner—sued AI music generator companies Suno and Udio. Additionally, ongoing cases from The New York Times target OpenAI and Microsoft.

This lawsuit marks the first over AI-generated text. As these rulings unfold, a pressing question arises: will determinations for one form of media extend to another?

John Strand, an IP and copyright attorney at Wolf Greenfield, stated, “The impact of copyrighted works on the market is increasingly vital in fair use analysis, and the book market has unique considerations compared to film.”

For Strand, the scenario concerning images seems to favor copyright holders since AI models are said to generate identical images based on their training data.

Even more startling revelations emerged from the AI verdicts. Companies have allegedly utilized 7 million pirated books to establish their AI training databases. To rectify this, they purchased physical copies, scanned them, and digitized the content. Unfortunately, the original owners of these 7 million physical books, which no longer have practical use, were left with destroyed copies. This operation involved buying the books, chopping them up, scanning their text, and then discarding them. According to ARS Technica, there are very few efficient methods for digitizing books, and they tend to be slow. The AI sector seems geared towards swift and disruptive approaches.

The destruction of millions of books illustrates the intense demand for content that AI companies require for their products.

AI and the Environment: Unfortunate Insights

Updates on Last Week’s Stories: Trump’s Phone

Two stories I reported last week have seen significant developments shortly thereafter.

The Trump-branded mobile phone, known as “T1,” has replaced its “America” pledge with “proudly America” and “vibrantly in America.” According to Barge.

Trump seems to be mirroring Apple’s strategy. While Apple navigates the manufacturing origin issues, it spotlights the American aspect of the iPhone by branding it as “designed in California.” What’s left unsaid is its assembly in China or India, along with components sourced from various countries. Trump and his family appear to have adopted a similarly ambiguous tagline, although their original commitments seem far more glaring.

The descriptor “American Proud Design” now featured prominently on Trump’s site appears to be an obvious nod to Apple’s branding.

Adhering to the “Made in the USA” label carries real legal implications. Companies face litigation over how many products are genuinely produced within the country, and major U.S. trade regulators have set standards for what constitutes that slogan. However, tracing a smartphone’s manufacturing history to meet these criteria proves to be quite complex, according to many experts.

While Trump aims to bring manufacturing back to America with his steep tariffs, it seems he has learned the lessons that other mobile companies have grappled with. Manufacturing smartphones solely in the U.S. is fraught with complications and limitations, creating significant challenges for the final product.

Catch up on last week’s Gold Trump Phone newsletter.

…and Online Age Verification

Last week, I discussed Smatty, a porn platform, returning to France. This week, the U.S. Supreme Court ruled in favor of the age verification checks mandated in Texas. Pornhub has blocked access for Texas residents for much of the past two years, protesting much like they did in France for three weeks.

Justice Clarence Thomas summarized the court’s rationale:

“HB 1181 simply requires adults to verify their age before accessing adult explicit materials,” Thomas stated in the majority opinion, which passed with a 6-3 ruling. “This law furthers the state’s significant interest in protecting children from sexually explicit content and appropriately allows users to verify their age using established forms of government-issued identification and shared transaction information.”

Justice Elena Kagan, along with two other liberal justices, voiced their dissent.

The ruling validates Texas laws and laws from nearly 20 other states implementing online age checks. The global climate seems to be shifting away from granting broader access to pornographic content under the guise of free speech rights.

Experts suggest that the flexible definition of obscenity under Texas law necessitates age checks on platforms containing adult-oriented materials.

“Today is disheartening for advocates of an open internet,” remarked GS Hans, a professor at Cornell Law School. “While the courts may not categorize this decision as a landmark ruling in this case, it fundamentally alters free speech jurisprudence and could establish encroachments on adult access by endorsing limitations on minor indecency.

We’ll monitor the situation closely in July when Pornhub intends to implement age checks in line with the Online Services Act.

Read more: A UK survey indicates that 8% of children aged 8 to 14 have encountered online pornography.

Explore More AI News

This Week in AI: WhatsApp Introduces Summary Feature and Nobel-Winning Genome Model

This new feature may seem minor, but even slight modifications to the globe’s most used messaging app can create a significant impact. Meta’s WhatsApp now provides AI-generated summaries of unread messages. According to Barge.

Apple previously experimented with message summaries—but that venture didn’t succeed, leading them to retract the feature. For companies known for strategically controlled launches, dropping the summary was quite an embarrassment. The difference here lies in Meta’s consistent track record of releasing AI products over the years.

In more AI-related news, I seldom find new technology captivating, but Google’s DeepMind AI Lab’s recent announcement appears promising for the healthcare sector. The new Alphagenome AI aims to offer comprehensive predictions regarding how a single mutation in human DNA can impact multiple biological processes governing genes. The developers of the Alphagenome previously won the Chemistry Prize for Alphafold, a program known for predicting protein structures.

This innovation raises compelling questions, potentially overtaking CRISPR, the groundbreaking technique regarding changes in humans when their genetic sequences are adjusted. The Alphagenome holds promise in shedding light on this enigmatic issue.

The Broader Tech Landscape