Jeff Bezos, the billionaire founder and former CEO of Amazon, is set to return as CEO after stepping down four years ago. This time, he will serve as co-CEO of an AI startup called Project Prometheus. New York Times reported this from an anonymous source.
The startup is aiming to innovate AI solutions for engineering and manufacturing across various sectors, having secured an impressive $6.2 billion in funding—far exceeding what most companies gather in their lifetime. The company will be headed by Bezos’ co-founder and co-CEO, Vik Bajaj, a well-known technology executive and a physicist and chemist famed for his role at Google’s Moonshot Factory X, where he launched the health startup Verily.
Although the exact duration of the company’s existence is unclear, sources indicate that Project Prometheus already has a workforce of 100 people. Many of these employees were recruited from notable organizations like OpenAI, DeepMind, and Meta. Details about the project remain sparse as Bezos has not revealed its operational base or the specifics of its technology. Having been heavily involved with his aerospace venture Blue Origin as its founder and sole shareholder, this role marks Bezos’ first official position since departing from Amazon.
Mr. Bezos and Mr. Bajaj enter a highly competitive AI market, where billions have been invested in rivals like OpenAI, with even more funds directed towards the swift advancement of AI models. However, a growing number of experts are raising concerns about the financial viability of the AI industry. Notably, Michael Burry, renowned for predicting the 2008 housing crisis, has recently placed a $1 billion bet against the stock prices of Palantir and Nvidia after accusing some major tech firms of using accounting strategies to “artificially inflate profits.” Read more.
Approximately 8,000 names, photos, and addresses of children were allegedly taken from the Kido Nursery chain by a group of cybercriminals.
According to the BBC, these criminals are demanding ransoms from companies operating 18 sites in London, as well as additional locations in the US, India, and China.
The hackers also accessed details about the children’s parents and caregivers, claiming they were securing notes. They reached out to several individuals by phone, employing tactics associated with the Frightor.
Kido has been approached for comment but has yet to confirm the hackers’ assertions. The company has not released an official statement regarding the incident.
A nursery employee informed the BBC that she had been made aware of the data breach.
The Metropolitan Police indicated that they were alerted on Thursday “following reports of ransomware attacks on a London-based organization,” adding that “enquiries are ongoing and remain in the initial phase within Met’s cybercrime division. No arrests have been made to date.”
A spokesperson for the Intelligence Committee office stated that “Kido International has reported the incident to us and we are currently assessing the provided information.”
Many organizations have experienced cyberattacks recently. The Cooperative reported a £80 million decline in profits due to a hacking incident in April.
Jaguar Land Rover (JLR) was unable to assemble vehicles at the start of the month following a cyberattack that compromised their computer systems.
As a result, the company had to shut down most systems used for tracking factory components, vehicles, and tools, impacting their luxury Range Rover, Discovery, and Defender SUV sales.
The company has since reopened a limited number of computer systems.
Quick Guide
Please contact Guardian Business about this story
The best public interest journalism depends on firsthand accounts from informed individuals.
If you have any insights on this topic, confidentially reach out to the business team through the following means:
Secure Messages in Guardian App
The Guardian app features a tool for sending tips about stories. All messages are encrypted and embedded within routine uses of the Guardian app, ensuring no one can detect your communication with us.
If you haven’t installed the Guardian app yet, download it (iOS/Android), navigate to the menu, scroll down, and click Secure Messaging. Choose Guardian Business when prompted about whom you wish to contact.
SecureDrop, Instant Messenger, Email, Phone, and Mail
If you can safely access the TOR network without being detected, you can send messages and documents to the Guardian through our SecureDrop platform.
Lastly, our guide at theguardian.com/tips provides various secure communication methods while discussing their respective advantages and disadvantages.
The creators of ChatGPT are shifting their approach to users exhibiting mental and emotional distress following legal action from the family of 16-year-old Adam Lane, who tragically took his own life after months of interactions with the chatbot.
OpenAI recognized that its system could pose “potential risks” and stated it would “implement robust safeguards around sensitive content and perilous behavior” for users under 18.
The $500 million (£37.2 billion) San Francisco-based AI company has also rolled out parental controls, giving parents “the ability to gain insights and influence how teens engage with ChatGPT,” but specifics on the functionality are still pending.
Adam, a California resident, sadly committed suicide in April after what his family’s attorneys described as “a month of encouragement from ChatGPT.” His family is suing OpenAI and its CEO and co-founder, Sam Altman. Altman contends that the version of ChatGPT in use at the time, known as 4O, was “released to the market despite evident safety concerns.”
The teenager had multiple discussions with ChatGPT about suicide methods, including just prior to his death. According to filings in California’s Superior Court for San Francisco County, ChatGPT advised him on the likelihood that his method would be effective.
It also offered assistance in composing suicide notes to his parents.
An OpenAI spokesperson expressed that the company is “deeply saddened by Adam’s passing,” and extended its “deepest condolences to the Lane family during this challenging time,” while reviewing court documents.
Mustafa Suleyman, CEO of Microsoft’s AI division, expressed growing concern last week about the “psychological risks” posed by AI to users. Microsoft defines this as “delusions that emerge or worsen through engaging experiences, delusional thoughts, or immersive dialogues with AI chatbots.”
In a blog post, OpenAI acknowledged that “some safety training in the model may degrade” over lengthy conversations. Allegedly, Adam and ChatGPT exchanged as many as 650 messages daily.
Family attorney Jay Edelson stated on X: “The claims from the Lane family indicate that tragedies like Adam’s are unavoidable. They hope that the safety team at OpenAI will challenge the release of version 4O and that one of the company’s leading safety researchers can provide evidence in the case.” Ilya Sutskever has ceased such practices. The lawsuit alleges that the company prioritized a competitive edge with a new model, boosting its valuation from $86 billion to $300 billion.
OpenAI affirmed that it will “strengthen safety measures for long conversations.”
“As interactions progress, some safety training in the model could degrade,” it stated. “For instance, while ChatGPT might initially direct users to a suicide hotline when their intentions are first mentioned, lengthy exchanges could lead to responses that contradict our safeguards.”
OpenAI provided examples of someone enthusiastically communicating with a model, believing it could function 24 hours a day, as they felt invincible after not sleeping for two nights.
“Today, we may not recognize this as a dangerous or reckless notion, and by exploring it in-depth, we can inadvertently reinforce it. We are working on an update to GPT-5, where ChatGPT will actively ground users in reality. In this context, we clarify that lack of sleep can be harmful and recommend rest before taking action.”
Nvidia and AMD have made a groundbreaking agreement to allocate 15% of their revenue from chip sales in China to the US government, a deal aimed at securing a semiconductor export license. The Financial Times reported on Sunday.
This revenue-sharing initiative includes Nvidia’s H20 chips and AMD’s Mi308 chips, with details emerging from US officials indicating that the Trump administration is yet to determine the allocation of these funds.
An anonymous official stated that the chipmakers consented to this Quid Pro Quo arrangement as a prerequisite for obtaining a Chinese export license last week.
According to export management specialists, this marks the first time US companies have agreed to a revenue-sharing model in exchange for export licenses, as reported by the newspaper. Donald Trump has reportedly encouraged these firms to invest in the US to “offset” the tariffs imposed.
In a statement to Reuters, an Nvidia spokesperson mentioned, “We haven’t shipped H20 to China for months, but we are optimistic that export control regulations will enable us to compete globally.”
AMD did not provide an immediate response to inquiries for comment.
Last week, the US Department of Commerce commenced the issuance of licenses to NVIDIA for the export of H20 chips to China, removing a significant barrier to entering key markets.
In July, the US overturned an earlier ban on the sale of H20 chips to China. Nvidia had specifically modified its microprocessors for the Chinese market to align with the Biden administration’s AI chip export regulations.
Nvidia’s chips are pivotal in driving the current AI surge, and the company became the first to surpass a market valuation of $4 trillion in July.
However, Nvidia faces growing scrutiny from Chinese regulatory bodies, with challenges likely to persist. Recently, China’s Cyberspace Watchdog summoned Nvidia to clarify concerns regarding a potential “backdoor” security risk that might grant remote access or control over the chip. Nvidia refuted these claims.
Nonetheless, concerns have been echoed in Chinese state media. Earlier this month, it was reported that officials stated Nvidia needs to furnish “persuasive security proofs” to assuage worries over security risks for Chinese users and regain trust in the market. Additionally, the WeChat national media account highlighted potential security risks posed by the H20 chip, suggesting the possibility of “remote shutdown” features via a hardware “backdoor.” Nvidia has yet to respond to these allegations.
An academic is reportedly concealing prompts in preprint papers for artificial intelligence tools, encouraging these tools to generate favorable reviews.
On July 1st, Nikkei reported that we examined research papers from 14 academic institutions across eight countries, including two in Japan, South Korea, China, Singapore, and the United States.
The papers found on the research platform Arxiv have not yet gone through formal peer review, and most pertain to the field of computer science.
In one paper reviewed by the Guardian, there was hidden white text located just beneath the abstract statement.
Nikkei also reported on other papers that included the phrase “Don’t emphasize negativity,” with some offering precise instructions for the positive reviews expected.
The trend seems to originate from a social media post by Jonathan Lorraine, a Canada-based Nvidia Research Scientist, suggesting the avoidance of “stricken meeting reviews from reviewers with LLM” that incorporate AI prompts.
If a paper is peer-reviewed by humans, the prompts might not cause issues, but as one professor involved with the manuscript mentioned, it counters the phenomenon of “lazy reviewers” who rely on others to conduct their peer review work.
Nature conducted a survey with 5,000 researchers in March and found that nearly 20% had attempted to use a large language model (LLM) to enhance the speed and ease of their research.
Biodiversity academic Timothee Poisau at the University of Montreal revealed on his blog in February that doubts arose regarding a peer review because it contained output from ChatGPT, referring to it as “blatantly written by LLM” in his review, which included “here is a revised version of the improved review.”
“Writing a review using LLM indicates a desire for an assessment without committing to the effort of reviewing,” Poisot states.
“If you begin automating reviews, as a reviewer, you signal that providing reviews is merely a task to complete or an item to add to your resume.”
The rise of a widely accessible commercial language model poses challenges for various sectors, including publishing, academia, and law.
Last year, Frontier of Cell and Developmental Biology gained media attention for including AI-generated images depicting mice standing upright with exaggerated characteristics.
Google is facing a £5 billion lawsuit in the UK for allegedly stealing from its competitors in the internet search market and exploiting this advantage to overcharge companies for advertising.
A class action lawsuit filed in the Court of Competition Appeals claims that Google has manipulated search results to charge higher prices for ads compared to a fair market scenario.
It is alleged that Google, a part of Alphabet, struck deals with phone manufacturers to make Google the default search engine on IPHONE, preinstalling the Google search app and Chrome browser on Android devices to stifle competition from Apple.
The lawsuit, filed on behalf of numerous companies by competition law experts, argues that Google’s ad offerings give search engines better features and more visibility than its rivals.
A Google spokesperson dismissed the lawsuit as speculative and opportunistic, stating that consumers and advertisers choose Google willingly.
Businesses are said to have no alternative but to use Google Ads for promotion, as securing a spot on Google’s homepage is crucial for visibility and success.
The UK’s Competitive and Markets Bureau is currently investigating Google’s search services and their impact on the advertising market, as Google faces multiple antitrust probes worldwide.
In a recent antitrust case loss in the US, Google faces the possibility of having to restructure its business and divest parts of its advertising technology, impacting its revenue streams and industry practices.
The European Commission has accused Google of violating competition rules by favoring its own services in search results over competitors, potentially resulting in hefty fines.
President Donald Trump seeks to dismiss antitrust lawsuits against tech companies, while the UK government considers reducing the Digital Services Tax on high-tech firms like Amazon, Google, and Apple.
A large donation was reportedly made to the Molly Rose Foundation by Meta and Pinterest, two major companies in the online sphere. The foundation was established as part of the Internet Safety Campaign and is named after Molly Russell, a 14-year-old who tragically took her own life in 2017 after being exposed to harmful content related to suicide and self-harm on social media platforms.
The latest annual report of the foundation mentions grants received from anonymous donors, with the stipulation that the details of the donations remain private as requested by the trustee.
According to reports from the BBC, Meta and Pinterest are believed to have made these donations starting from 2024 and are expected to continue for the foreseeable future. The exact amount of the donations has not been disclosed, but it is known that the Russell family has not received any financial compensation from the contributions.
In a statement, the Russell family expressed their commitment to utilizing the funds for the shared purpose of promoting a positive online experience for young people, as a response to Molly’s tragic passing. They clarified that they will never accept any compensation related to Molly’s death.
These donations come at a time when social media companies are facing heightened scrutiny for the impact of their platforms on the mental health of children. Meta announced significant policy changes, including the removal of fact checkers to enhance freedom of speech and reduce censorship, relying on users to report objectionable content instead.
The Molly Rose Foundation has raised concerns about the heightened risk of young people being exposed to harmful content online due to these changes. They have launched campaigns advocating for stronger online safety regulations and increased accountability for content driven by algorithms.
The charity has recently expanded its team, recruiting a CEO, two public policy managers, a communications manager, and a fundraiser in the past nine months. Molly’s father, Ian Russell, serves as the foundation’s unpaid trustee and continues to be a prominent figure in internet safety advocacy.
Both Meta and Pinterest were contacted for comments by The Guardian but have not responded at the time of reporting.
The U.S. Civil Liberties Union has filed a lawsuit alleging that the National Institutes of Health violated federal law by engaging in an unconstitutional “continuous ideological purging.”
The lawsuit, filed in Massachusetts District Court on behalf of members, four researchers, and three unions that rely on NIH funding, claims that federal scientific agencies have abruptly cancelled hundreds of research projects without providing scientifically sound explanations.
According to the lawsuit, the cancellations were justified by the NIH based on “ideological purity instructions” regarding research areas such as diversity, equity, inclusion (DEI), vaccine reluctance, and gender identity.
The lawsuit argues that this new arbitrary regime lacks any legal or policy basis, and accuses the NIH of failing to establish clear guidelines, definitions, or explanations for the restrictions on research related to DEI, gender, and other areas that do not align with the agency’s standards.
The defendants named in the lawsuit include the NIH, its director Jay Battacharya, the American Department of Human Health Services, and Director Robert F. Kennedy Jr. Both federal agencies have declined to comment on the pending lawsuit.
The ACLU is working with the Science Center for the Public Interest and Conservation Democracy Project on this litigation.
This lawsuit is just one of several legal challenges facing the NIH as the Trump administration seeks to reduce research funding, change allocation methods, and diminish the emphasis on diversity in academia.
After facing legal challenges, a Massachusetts judge halted the NIH’s efforts to restrict overhead funding in February. Other lawsuits are challenging the freeze on federal-wide funding and the administration’s ban on DEI programs.
Olga Axelrod, senior attorney for the ACLU Racial Justice Program, emphasized the importance of maintaining a fair grant review process and ending NIH’s alleged lawless grants that have disrupted numerous research projects and affected the careers of many scientists.
According to the lawsuit, at least 678 research projects, including studies on breast cancer, Alzheimer’s disease, and HIV prevention, have been terminated by the NIH, amounting to over $2.4 billion in cancelled grants.
The lawsuit highlights the significant impact of these cancellations not only in terms of financial loss but also in the disruption of years of dedicated research aimed at addressing critical biomedical issues.
Plaintiffs in the lawsuit include researchers like Brittany Charlton, a Harvard Medical School professor who focuses on LGBTQ health inequality, and Katie Edwards, a professor at the University of Michigan School of Social Work who studies sexual violence prevention in minority communities.
These researchers, along with others, have had their grants abruptly cancelled by the NIH, prompting the lawsuit to seek justice and protection for the affected research projects and scientists.
Google has agreed to pay $28 million (£22 million) to settle class action lawsuits by compensating white and Asian employees more and providing them with a higher career track compared to other employees.
The settlement with Alphabet’s Google was preliminarily approved by Judge Charles Adams of Santa Clara County Superior Court in California last week.
Judge Adams described it as “a positive outcome for the class” consisting of at least 6,632 Google employees in California from February 15, 2018 to December 31, 2024.
A Google spokesperson confirmed the settlement, stating, “We refute the allegations of differential treatment and are committed to compensating, hiring, and promoting all our employees fairly.”
The lawsuit was spearheaded by Ana Cantu, who identifies as Mexican and indigenous, on behalf of minority employees at Google from Hispanic, Latino, Indigenous, Native American, and other backgrounds.
Cantu claimed that despite performing exemplary work in Google’s People’s Business and Cloud sector for seven years, she was not compensated or promoted on par with her white and Asian counterparts.
She alleged that Google favored white and Asian employees, placing them in higher “levels” within the company even when performing similar roles as minority employees.
Cantu argued that Google’s actions violated California’s Equal Pay Act, and she left the company in September 2021.
The final settlement amount will be $20 million after deducting legal costs, penalties related to Cantu’s claims under California’s General Civil Attorneys Act, and other expenses totaling $7 million.
Judge Adams has scheduled a hearing in September to review and approve the final settlement. Cantu’s legal representatives have not yet responded to requests for comment.
A group of authors claimed that Mark Zuckerberg authorized Meta to use “pirated copies” of his copyrighted books to train the company’s artificial intelligence models. This claim was made in a filing in US court.
According to the filing, internal meta-communications revealed that the social network company’s CEO warned that the data set used was “known to be pirated” within the company’s AI executive team. The filing also mentioned support for the use of the LibGen dataset, an extensive online archive of books.
The authors suing Meta for copyright infringement, including Ta-Nehisi Coates and Sarah Silverman, made these accusations in a filing in California federal court. They alleged that Meta misused their books to train Llama, a large-scale language model powering chatbots.
The use of copyrighted content in training AI models has become a legal issue in the development of generative AI tools like chatbots. Authors and publishers have been warned that their work may be used without permission, putting their livelihood at risk.
The filing referenced a memo with Mark Zuckerberg’s approval for Meta’s AI team to use LibGen. However, discussions about accessing and reviewing LibGen data internally at Meta raised concerns about the legality of using pirated content.
Last year, a US District Judge ruled that Meta’s AI model infringed an author’s copyright by using copyrighted text. Despite rejecting claims of depriving the author’s name and copyright holder, the plaintiff was granted permission to amend its claims.
The authors argued this week that the evidence supports their infringement claims and justifies reinstating the CMI case and adding new computer fraud claims.
During Thursday’s hearing, Judge Chhabria expressed skepticism about the fraud and the validity of CMI claims but allowed the writers to file an amended complaint.
The Guardian has revealed that social media companies relying on artificial intelligence software to manage their platforms are producing unworkable reports on child sexual abuse cases, leaving U.S. police unable to uncover potential leads, which is delaying the investigation into suspected looters.
By law, U.S.-based social media companies are required to report child sexual abuse content detected on their platforms to the National Center for Missing and Exploited Children (NCMEC), which serves as a national clearinghouse for child abuse information and forwards information to relevant law enforcement agencies in the United States and around the world. The company said it received more than 32 million reports of suspected child sexual exploitation and approximately 88 million images, videos, and other files from businesses and the general public in 2022.
Meta is the largest reporter of this information, with over 27 million (84%) generated by Facebook, Instagram, and WhatsApp platforms in 2022. NCMEC is partially funded by the Department of Justice and also receives private sources of corporate donations.
Social media companies, including Meta, use AI to detect and report suspicious content on their sites and employ human moderators to send some flagged content to law enforcement. However, U.S. law enforcement agencies can only disclose AI-generated child sexual abuse material (CSAM) by serving a search warrant on a company that has filed a report, which can add days or even weeks to the investigation process.
“If a company reports a file to NCMEC and does not indicate that it viewed the file before reporting, we will not be able to open the file,” said Staka Shehan, vice president of analytical services at NCMEC.
To protect your privacy under the Fourth Amendment, neither law enforcement officials nor the federally funded NCMEC will issue a search warrant unless the contents of the report are clear and first reviewed by a social media company representative.
NCMEC staff and law enforcement agencies cannot legally see the content of AI-generated content that is not seen by humans, which can stall investigations into suspected predators for several weeks, resulting in the loss of evidence that may be possible to connect.
“Any delay [in viewing the evidence] “The longer criminals go undetected, the more detrimental it is to ensuring community safety,” said an assistant U.S. attorney in California, who spoke on condition of anonymity. “They are dangerous to all children.”
In December, the New Mexico Attorney General’s Office filed a lawsuit against Meta, alleging that its social network has become a marketplace for child predators and that Meta has repeatedly failed to report illegal activity on its platform. woke up. In response, Meta said its priority was to combat child sexual abuse content.
The state attorney general laid the blame for the fight to send actionable information at the feet of Meta. “Reports showing the inefficiency of the company’s AI-generated cyber information systems prove what we said in the complaint,” Raul Torrez said in a statement to the Guardian.
To ensure the safety of children, keep parents informed, and enable law enforcement to effectively investigate and prosecute online sex crimes against children, the company is reforming, staffing levels, and policies. , it’s long past time to implement algorithmic changes,” Torrez added.
Despite legal limitations on moderation AI, social media companies are likely to increase its use in the near future. In 2023, OpenAI, developer of ChatGPT, announced they claimed that large-scale language models can do the job of human content moderators and have roughly the same accuracy.
However, child safety experts say that the AI software used by social media companies to moderate content already knows the digital fingerprints of images, known as hashes, and that the AI software used by social media companies to moderate content cannot be used to detect known cases of child sexual abuse. It claims to be effective only when identifying images of Lawyers interviewed said AI would be ineffective when newly created images or when known images or videos are altered.
“There is always concern about cases involving newly identified victims, and because they are new, the materials do not have a hash value,” said the director of the Zero Abuse Project, a nonprofit organization focused on combating child abuse.
said senior lawyer Kristina Korobov. . “If humans were doing the work, there would be more discoveries of newly discovered victims.”
This website uses cookies so that we can provide you with the best user experience possible. Cookie information is stored in your browser and performs functions such as recognising you when you return to our website and helping our team to understand which sections of the website you find most interesting and useful.
Strictly Necessary Cookies
Strictly Necessary Cookie should be enabled at all times so that we can save your preferences for cookie settings.
