Hub AI and society

Artificial Intelligence is a key technology that can help us solve major societal challenges such as climate change, energy supply, healthcare, education or logistics. AI can solve specific problems more effectively by supporting us in decision-making, automating solutions and thus scaling them, or discovering completely new solutions. But the use of AI also poses new risks, for example when it comes to surveillance or questions of social justice.

What is our society doing with AI – and what is AI doing to our society? We shed light on this question in our AI and Society Content Hub.

AI and society

Oct 28, 2025Oct 28, 2025

OpenAI restructures under new foundation, Microsoft takes 27 percent stake

AI and society

Oct 19, 2025

The long-predicted deepfake dystopia has arrived with Sora 2

AI and society

Aug 23, 2025Aug 23, 2025

Google downplays AI's environmental impact in new study

AI and society

Dec 4, 2025Dec 4, 2025

Short

The European Union is planning a major expansion of its AI infrastructure. The European Investment Bank (EIB) and the European Commission want to build up to five AI gigafactories across Europe to boost compute capacity for advanced AI models and reduce the region's dependence on foreign technology.

The Commission plans to fund the effort with 20 billion euros through its InvestAI program, and the EIB is considering additional loans. Each site will include about 100,000 high-performance AI chips, described as "the most advanced" available and roughly four times more than existing facilities.

"AI gigafactories will train the most complex, very large AI models, which require extensive computing infrastructure for breakthroughs in domains such as medicine, cleantech and space."

The project falls under the EIB's TechEU program, which aims to mobilize 250 billion euros in investment by 2027.

EIB

AI and society

Dec 4, 2025Dec 4, 2025

Short

Nvidia and OpenAI have not yet signed their planned 100 billion dollar deal. Nvidia CFO Colette Kress confirmed this on Tuesday during a conference in Arizona. Even though both companies announced plans in September to provide 10 gigawatts of Nvidia systems for OpenAI, the arrangement is still only a memorandum of understanding. Kress said the two sides are still working toward a final agreement.

The holdup raises new questions about the circular business structures that have become common in the tech industry, where large companies invest in startups that then spend the money on the investor's own products. Any future revenue from the OpenAI deal is not included in Nvidia's current 500 billion dollar forecast. A separate 10 billion dollar investment in competitor Anthropic also remains pending.

Reuters

AI and society

Dec 3, 2025Dec 3, 2025

Short

A new study from MATS and Anthropic shows that advanced AI models like Claude Opus 4.5, Sonnet 4.5, and GPT-5 can spot and exploit smart contract vulnerabilities in controlled tests. Using the SCONE-bench benchmark, which includes 405 real smart contract exploits from 2020 to 2025, the models produced simulated damage of up to 4.6 million dollars.

In a separate experiment, AI agents reviewed 2,849 new contracts and uncovered two previously unknown vulnerabilities. GPT-5 generated simulated revenue of 3,694 dollars at an estimated API cost of about 3,476 dollars, averaging a net gain of 109 dollars per exploit. All experiments were run in isolated sandbox environments.

The researchers say the findings point to real security risks but also show how the same models could help build stronger defensive tools. Anthropic recently released a study suggesting that AI systems can play a meaningful role in improving cybersecurity.

Anthropic

AI and society

Dec 3, 2025Dec 3, 2025

Short

Google is testing an AI feature in its "Discover" news feed that automatically rewrites editorial headlines, often turning them into shorter, more provocative, or simply incorrect versions. The result is the kind of engagement-focused headline the company warns against in its own Discover rules, which explicitly reject clickbait.

Google's guidelines set clear limits on the kinds of headlines Discover should surface. | Image: via Google

Some of the changes are striking. At Ars Technica, the headline "Valve's Steam Machine looks like a console, but don't expect it to be priced like one" was replaced with the inaccurate "Steam Machine price revealed." A Mindfactory report titled "Radeon RX 9070 XT Outsells The Entire NVIDIA RTX 50 Series On Popular German Retailer" was shortened to "AMD GPU tops Nvidia."

Google told The Verge that the feature is a small test meant to help users grasp articles more quickly. Still, the AI-generated rewrites replace newsroom decisions with Google's own automated framing, raising concerns that the company is using AI to tighten its grip on how news is shaped and delivered on top of the influence it already exercises.

The Verge

AI and society

Dec 3, 2025Dec 3, 2025

Anthropic prepares for a potential IPO race with OpenAI

AI and society

Dec 1, 2025Dec 1, 2025

Copyright pressure mounts as OpenAI battles over newspapers and pirate libraries

AI and society

Dec 1, 2025Dec 1, 2025

New York targets algorithmic pricing with first-in-the-US disclosure law

AI and society

Nov 28, 2025Nov 28, 2025

Short

A new Financial Times analysis found that several OpenAI partners have taken on roughly 96 billion dollars in debt to fund data center and chip expansion. The list includes major players like Oracle and Softbank, along with specialized providers such as Coreweave. According to the report, Coreweave's liabilities and lease obligations far exceed its expected annual revenue of five billion dollars.

The borrowing isn't limited to a few companies. Bank of America says the five largest tech firms, including Amazon and Microsoft, have issued 121 billion dollars in new debt this year - about four times their usual annual average. Deutsche Bank also reports rising credit default swap costs for companies like Oracle, pointing to growing concern among lenders and investors.

AI and society

Nov 27, 2025Nov 27, 2025

Short

OpenAI has rejected the lawsuit filed by the family of 16-year-old Adam Raine, saying the company is not responsible for the teenager's suicide. In a court filing on Tuesday, the company argued that Raine misused ChatGPT in ways that violated its terms of use and intentionally worked around the system's safety filters. According to OpenAI's lawyers, the model pointed him to support resources more than 100 times, but Raine repeatedly disguised his intentions to bypass those warnings. The family and their attorney, Jay Edelson, claim the situation unfolded differently, accusing the company of releasing the heavily criticized GPT-4o model despite concerns about flattery and safety. One OpenAI developer recently described the model as "insufficiently aligned" when interacting with users.

In a blog post, OpenAI said it intends to handle the case respectfully but will have to reveal "difficult facts" about Raine's mental health as part of its defense. The company says the plaintiffs cited only selected chat excerpts, prompting OpenAI to submit the full, sealed transcripts to the court.

OpenAI NBC News