OpenAI announced three scientific advances in mathematics, physics and computer science on Monday, saying they were achieved with the help of its artificial intelligence (AI) models. However, the ChatGPT developer has faced criticism over how it communicates its findings to the scientific community and its role in verifying its own results.
In an article published Tuesday, U.S. technology news outlet The Verge quoted two researchers, Constantin Kogler of the Institute for Advanced Study and Andreas Karch, a physics professor at the University of Texas at Austin, who raised concerns about the company’s approach.
According to OpenAI, the company has demonstrated the local smoothing conjecture in mathematics, an open problem since 1991, as well as two related conjectures: the Kakeya conjecture and the Stein restriction conjecture. The findings represent progress in the mathematical study of waves.
In theoretical computer science, OpenAI said it had refuted the Exponential Time Hypothesis (ETH), a fundamental conjecture stating that certain computational problems cannot be solved efficiently, even with optimized algorithms.
In physics, the company said it had produced a rigorous mathematical proof of the existence of certain four-dimensional quantum field theories, including quantum chromodynamics (QCD), which describes the strong interaction between elementary particles, and Yang-Mills theories.
Read also:
After their son’s suicide, Charbel and Soha Habib’s fight against uncontrolled AI
In early September, OpenAI also said it had solved the century-old Navier-Stokes problem in 88 hours. The equations describe the movement of fluids, including gases and liquids, and remain a major challenge in mathematics.
Five complaints filed in Europe against Scale AI
The Austrian digital rights organization Noyb announced that it had filed five complaints in Europe against U.S. company Scale AI, accusing it of illegally collecting workers’ personal data through its AI training platform, Outlier.
Two complaints were filed in Austria, with one each in France, the Netherlands and Italy, the organization said.
Outlier, which says it works with more than 900,000 independent contractors tasked with evaluating and correcting AI-generated responses, allegedly installs trackers on workers’ devices, allowing companies including TikTok, Facebook, LinkedIn, Sprig and Datadog to collect personal information
The trackers reportedly transmit information every few seconds, including IP addresses, contact details, unique identifiers and data about the pages users visit. This information could then be used for targeted advertising or profiling, the organization said.
“In Europe, this widespread workplace surveillance clearly violates workers’ fundamental right to data protection,” Noyb said.
The organization also criticized the repeated use of biometric verification procedures operated by U.S. company Persona, which involve recording workers’ faces on video and collecting their identity documents.
Meta seeks $97.5 billion to finance AI expansion
Meta Platforms, the parent company of Facebook, Instagram and WhatsApp, has announced a share offering of up to $97.5 billion, its largest to date, to accelerate the development of its AI infrastructure.
Chipmakers Nvidia, Advanced Micro Devices (AMD) and Broadcom have expressed interest in the financing plan and could invest up to $5 billion combined. Nvidia and AMD have separately announced plans to invest $3 billion and $1 billion, respectively.
The involvement of chipmakers in financing one of their major customers has raised concerns about increasingly circular capital flows in the AI industry. Meta is investing heavily in chips and data centers as it seeks to compete with OpenAI and Anthropic.
The fundraising reflects a broader trend among technology companies, which are increasingly turning to bond markets and share offerings to finance AI infrastructure rather than relying primarily on their cash reserves.
Alphabet, Google’s parent company, raised $84.75 billion in equity in June. On Monday, SpaceX announced that an affiliate of Apollo Global Management would purchase AI chips and computing equipment and lease them back to Elon Musk’s company under an agreement involving approximately $40 billion in principal payments.
Meanwhile, the Paris Court of Appeal rejected Meta’s appeals against measures imposed by France’s competition authority requiring the company to “negotiate in good faith” with French media outlets over the use of their content, the company announced Thursday.
New York restricts minors’ access to AI companions
New York Gov. Kathy Hochul signed a bill into law Wednesday restricting minors’ interactions with AI companions, adopting a definition broad enough to cover most generative AI chatbots on the market.
Several U.S. states, including California, have already adopted laws regulating minors’ use of AI companions, but New York’s legislation is among the broadest in scope.
Tech:NYC, a New York-based technology industry association, criticized the legislation, particularly its “vague language,” which it said “raises serious questions” about whether the law violates the First Amendment to the U.S. Constitution, which protects freedom of speech.
Read also:
When Internet users use AI to make the Hamas-Israel war more aesthetic
The law requires chatbot operators to verify users’ ages and restrict certain AI features when a user is identified as a minor.
Chatbots must not generate content suggesting that they are “human, alive or experiencing human emotions.” They must also refrain from presenting interactions with minors as personal or professional relationships or assuming the role of a mentor.
The law further prohibits chatbots from presenting generated text as a “personal opinion” or asking minors questions about their feelings or psychological state.
Finnish police investigate Google data center project
Finnish police have opened a criminal investigation into the environmental impact of preparatory work at a site designated for a Google data center.
In September, the U.S. technology giant announced plans to invest €13 billion ($15 billion) in digital infrastructure across four sites in Finland over the next two years as part of its efforts to expand its AI capabilities.
The Oulu police department said it had opened an investigation into Finnish company Tuike, which represents Google, over work carried out in Muhos, in northern Finland. The case is being investigated as suspected environmental degradation, police said, adding that further potential offenses could be identified as the investigation progresses.
Preliminary findings pointed to “ambiguities and deficiencies in the drainage and water management at the construction site of the Tuike Finland Oy data center project,” according to police.
The investigation was launched following a request from Finland’s Supervisory Authority (LVV).
Police said the investigation was expected to take considerable time and conclude in 2027, with a further update likely in December.
![Minicor Raises $3.8M Seed for AI Agents [2026]](https://tech-insider.org/wp-content/uploads/2026/10/minicor-3-8m-seed-ai-agents-legacy-desktop-software-2026-gen.webp "Minicor Raises $3.8M Seed for AI Agents [2026]")
