By Kaushik Sridasyam in This Week in AI —

Billion-Dollar Drama, Broken Promises, and a Courtroom Showdown

The world of AI never has a dull week. From tech titans throwing around $97 billion offers (with a catch) to a tech giant quietly backtracking on an AI promise, and even a landmark courtroom clash over AI’s training data – it’s been a rollercoaster.

Billion-Dollar Drama, Broken Promises, and a Courtroom Showdown

The world of AI never has a dull week. From tech titans throwing around $97 billion offers (with a catch) to a tech giant quietly backtracking on an AI promise, and even a landmark courtroom clash over AI’s training data – it’s been a rollercoaster.

Elon Musk’s $97 Billion OpenAI Power Play (Yes, Billion with a B)

It wouldn’t be an AI news roundup without a little Elon Musk drama. Last week, Musk made headlines by dangling an eye-popping $97.4 billion offer to take control of OpenAI – the company behind ChatGPT – but (in true Musk fashion) there was a catch. He basically said “I’ll buy you... unless you do what I want without me buying you.” More specifically, Musk’s group announced it would withdraw the $97 billion bid if OpenAI’s board agreed to abandon plans to become a for-profit company. In other words, Musk is telling OpenAI: stick to your original non-profit mission and I won’t need to buy you.

Why all this melodrama? Some history: Musk co-founded OpenAI in 2015 but quit the board in 2018 over disagreements about its direction. OpenAI started as a nonprofit lab, but later shifted to a capped-profit model to secure billions in investments (hello, Microsoft). Musk has openly accused OpenAI of “betraying” its original mission and becoming too commercial. So this massive bid was his power move to “set things right” – or at least spark a conversation about it. OpenAI’s response? A polite “no thanks.” In fact, OpenAI’s board hadn’t even received a formal offer, and insiders shrugged it off as an unsolicited, somewhat theatrical proposal. CEO Sam Altman even quipped that he’d rather spend one-tenth of that money to buy X (formerly Twitter) from Musk instead – a sarcastic burn that underscores the feud. Musk fired back by calling Altman a “swindler,” proving that even AI governance can turn into a soap opera.

Why it matters: This tussle isn’t just billionaire ego sparring – it raises real questions about the soul of OpenAI and the future of AI governance. Should advanced AI be guided by altruistic, non-profit ideals, or do we accept that big bucks and corporate control are inevitable for rapid innovation? Musk’s stunt highlights the tension between keeping AI research open for humanity’s benefit and the huge commercial interests now at play. It’s a drama with hefty stakes (and price tags), so stay tuned for the next episode in the Musk vs. OpenAI saga.

Google Quietly Ditches Its “No AI Weapons” Pledge – Are We the Baddies Now?

Remember when Google said it wouldn’t let its AI tech be used for weapons or surveillance, to “not be evil”? Well… about that. Last week we learned Google quietly scrubbed its famous pledge against weaponized AI from its official guidelines. In an all-hands meeting, execs confirmed that Google’s new AI ethics policy no longer forbids using AI for building weapons or surveillance tools – a stark reversal from the principled stance it took in 2018 after employee protests. They also announced they’re nixing some diversity and inclusion (DEI) initiatives for good measure. Cue the internal outrage: Googlers flooded the company message board with memes and posts expressing dismay. One popular meme had CEO Sundar Pichai googling “how to become a weapons contractor?”, while another referenced a comedy sketch with the line: “Are we the baddies?”. Yet another showed Sheldon from The Big Bang Theory asking why Google would drop its red line on weapons – and then answering himself, “Oh, that’s why,” alluding to Google cozying up to defense contracts. Ouch.

Google’s leadership tried to justify the about-face. In a blog post, DeepMind CEO Demis Hassabis and VP James Manyika described an “increasingly complex geopolitical landscape” and said it’s important for companies and governments to work together in the interest of “national security”. Reading between the lines: Google doesn’t want to be left out of big military AI projects (especially as rivals like Amazon and Microsoft are signing lucrative defense deals). A spokesperson pointed to the blog, emphasizing that democratic countries should lead in AI development and that working with governments can help “protect people” and “support national security”.

Still, the optics are wild. Google went from “we won’t even touch military AI” to “well, if it’s for national security… ¯\(ツ)/¯”. Internally, many employees are uneasy about this shift – after all, it was their revolt during the Pentagon’s Project Maven that led to the original pledge. Externally, ethicists worry this could erode trust in Big Tech’s self-imposed limits. Google dropping its no-weapons vow is a big sign of how competitive and strategic the AI race has become, where even past promises may not stand in the way of “the future.” Whether you view it as pragmatic or problematic, it’s a landmark moment in AI ethics (and one that inspired killer memes, literally).

AI has been gobbling up text and images from the internet like there’s no tomorrow – but now the courts are starting to ask, “Hey, is that even legal?” Last week, we got the first major U.S. court ruling on AI and copyright, and it’s a doozy. In a lawsuit by publishing giant Thomson Reuters against a now-defunct AI startup called Ross Intelligence, a federal judge made it clear that you can’t just copy someone’s copyrighted text to train your AI and call it fair use. Ross had used content from Thomson Reuters’ Westlaw (a massive database of legal documents) to train an AI lawyer of sorts – essentially aiming to answer legal questions without lawyers needing Westlaw. Thomson Reuters sued, claiming “Hey, you stole our stuff to build a competing product!” And the judge agreed. U.S. Circuit Judge Stephanos Bibas ruled that Ross infringed copyright and couldn’t lean on the “fair use” excuse for wholesale copying. Notably, he pointed out that Ross “meant to compete with Westlaw by developing a market substitute” – basically replacing the original instead of transforming it – which torpedoed their fair use defense. In plain English: if your AI is trained on someone else’s material and tries to serve the same purpose, that’s not gonna fly.

This decision is huge because it’s the first time a court has weighed in on the blurry practice of AI data scraping. For AI companies, it’s a wake-up call that the Wild West days of training on whatever data you can grab might be coming to an end. For content creators – authors, artists, news outlets – it’s a welcome precedent. In fact, observers say this win bolsters the case for artists and writers who don’t want generative AIs devouring their work without permission or compensation. (You could practically hear the cheers from publishing houses when the ruling came down.) Of course, this is just one case, and bigger battles are on the horizon – there are other lawsuits brewing against AI firms for using copyrighted books and code. But as a first shot fired, Judge Bibas’s ruling sends a clear message: AI needs to play by the existing copyright rules. Training a clever algorithm doesn’t magically exempt you from the law. Going forward, AI developers might have to get more creative (or more licensed) with their training data, unless they want to tango with lawyers.

Conclusion: AI Moves Fast – Buckle Up!

From corporate boardrooms to courtrooms, last week’s AI news had a bit of everything: billionaire power moves, tech giants eating their words, and legal lines in the sand. It’s a sign of a maturing field – the stakes (and the $$$) are getting bigger, the ethical dilemmas thornier, and the rules are now being written in real time. For the general public, it’s equal parts fascinating and perplexing. One minute you’re playing with a fun chatbot; the next, you’re hearing about dueling CEOs and AI law precedents.

The key takeaway? AI is not just about technology – it’s about people, power, and principles. We’re seeing visionaries and companies grapple with how AI should evolve: Who controls it? On what terms? With what moral guardrails? Last week gave us some hints (and a few juicy conflicts) that will shape the answers. And if you found all this a bit overwhelming, don’t worry – you’re not alone. The AI world may be complex, but we’ll keep breaking it down with a wink and a smile. Stick around for the next round of AI news – if it’s anything like last week, you won’t want to miss it!

Sources: