Big Tech News Articles
A US federal appeals court ruled last week that so-called geofence warrants violate the Fourth Amendment’s protections against unreasonable searches and seizures. Geofence warrants allow police to demand that companies such as Google turn over a list of every device that appeared at a certain location at a certain time. The US Fifth Circuit Court of Appeals ruled on August 9 that geofence warrants are “categorically prohibited by the Fourth Amendment” because “they never include a specific user to be identified, only a temporal and geographic location where any given user may turn up post-search.” In other words, they’re the unconstitutional fishing expedition that privacy and civil liberties advocates have long asserted they are. Google ... is the most frequent target of geofence warrants, vowed late last year that it was changing how it stores location data in such a way that geofence warrants may no longer return the data they once did. Legally, however, the issue is far from settled: The Fifth Circuit decision applies only to law enforcement activity in Louisiana, Mississippi, and Texas. Plus, because of weak US privacy laws, police can simply purchase the data and skip the pesky warrant process altogether. As for the appellants in the case heard by the Fifth Circuit, well, they’re no better off: The court found that the police used the geofence warrant in “good faith” when it was issued in 2018, so they can still use the evidence they obtained.
Note: Read more about the rise of geofence warrants and its threat to privacy rights. For more along these lines, see concise summaries of deeply revealing news articles on Big Tech and the disappearance of privacy from reliable major media sources.
OnlyFans makes reassuring promises to the public: It’s strictly adults-only, with sophisticated measures to monitor every user, vet all content and swiftly remove and report any child sexual abuse material. Reuters documented 30 complaints in U.S. police and court records that child sexual abuse material appeared on the site between December 2019 and June 2024. The case files examined by the news organization cited more than 200 explicit videos and images of kids, including some adults having oral sex with toddlers. In one case, multiple videos of a minor remained on OnlyFans for more than a year, according to a child exploitation investigator who found them while assisting Reuters. OnlyFans “presents itself as a platform that provides unrivaled access to influencers, celebrities and models,” said Elly Hanson, a clinical psychologist and researcher who focuses on preventing sexual abuse and reducing its impact. “This is an attractive mix to many teens, who are pulled into its world of commodified sex, unprepared for what this entails.” In 2021 ... 102 Republican and Democratic members of the U.S. House of Representatives called on the Justice Department to investigate child sexual abuse on OnlyFans. The Justice Department told the lawmakers three months later that it couldn’t confirm or deny it was investigating OnlyFans. Contacted recently, a department spokesperson declined to comment further.
Note: For more along these lines, see concise summaries of deeply revealing news articles on sexual abuse scandals from reliable major media sources.
More than 300 million children across the globe are victims of online sexual exploitation and abuse each year, research suggests. In what is believed to be the first global estimate of the scale of the crisis, researchers at the University of Edinburgh found that 12.6% of the world’s children have been victims of nonconsensual talking, sharing and exposure to sexual images and video in the past year, equivalent to about 302 million young people. A similar proportion – 12.5% – had been subject to online solicitation, such as unwanted sexual talk that can include sexting, sexual questions and sexual act requests by adults or other youths. Offences can also take the form of “sextortion”, where predators demand money from victims to keep images private, and abuse of AI deepfake technology. The US is a particularly high-risk area. The university’s Childlight initiative – which aims to understand the prevalence of child abuse – includes a new global index, which found that one in nine men in the US (equivalent to almost 14 million) admitted online offending against children at some point. Surveys found 7% of British men, equivalent to 1.8 million, admitted the same. The research also found many men admitted they would seek to commit physical sexual offences against children if they thought it would be kept secret. Child abuse material is so prevalent that files are on average reported to watchdog and policing organisations once every second.
Note: New Mexico's attorney general has called Meta the world's "single largest marketplace for paedophiles." For more along these lines, read our concise summaries of news articles on Big Tech and sexual abuse scandals.
[Tim] Berners-Lee, a British computer scientist, [came] up with the idea for a “world wide web” as a way of locating and accessing documents that were scattered all over the internet. He was able to do this because the internet, which had been publicly available since January 1983, enabled it. The network had no central ownership or controller. The result was an extraordinary explosion of creativity, and the emergence of ... a kind of global commons. However, the next generation of innovators to benefit from this freedom – Google, Facebook, Amazon, Microsoft, Apple et al – saw no reason to extend it to anyone else. The creative commons of the internet has been gradually and inexorably enclosed. Google and Apple’s browsers have nearly 85% of the world market share. Microsoft and Apple’s two desktop operating systems have almost 90%. Google runs about 90% of global search. More than half of all phones come from Apple and Samsung, while 99% of mobile operating systems are from Google or Apple. Apple and Google’s email clients manage nearly 90% of global email. GoDaddy and Cloudflare serve about 50% of global domain name system requests. And so on. One of the consequences of this concentration, say Farrell and Berjon, is that the creative possibilities of permissionless innovation have become increasingly constrained. The internet has become an extractive and fragile monoculture. We can revitalise it, but only by “rewilding” it.
Note: For more along these lines, see concise summaries of deeply revealing news articles on corporate corruption from reliable major media sources.
A federal appeals court on Tuesday refused to hold five major technology companies liable over their alleged support for the use of child labor in cobalt mining operations in the Democratic Republic of the Congo. In a 3-0 decision, the U.S. Court of Appeals for the District of Columbia ruled in favor of Google parent Alphabet, Apple, Dell Technologies, Microsoft and Tesla, rejecting an appeal by former child miners and their representatives. The plaintiffs accused the five companies of joining suppliers in a "forced labor" venture by purchasing cobalt, which is used to make lithium-ion batteries. Nearly two-thirds of the world's cobalt comes from the DRC. According to the complaint, the companies "deliberately obscured" their dependence on child labor, including many children pressured into work by hunger and extreme poverty, to ensure their growing need for the metal would be met. The 16 plaintiffs included representatives of five children who were killed in cobalt mining operations. Circuit Judge Neomi Rao said the plaintiffs had legal standing to seek damages, but did not show the five companies had anything more than a buyer-seller relationship with suppliers. Terry Collingsworth, a lawyer for the plaintiffs ... said his clients may appeal further. The decision provides "a strong incentive to avoid any transparency with their suppliers, even as they promise the public they have 'zero tolerance' policies against child labor," he said. "We are far from finished seeking accountability."
Note: Unreported deaths of children, devastating diseases, toxic environments, and sexual assault are just some of the tragedies within the hidden world of cobalt mining in the DRC. Furthermore, entire communities have been forced to leave their homes to make way for new mining operations. For more along these lines, see concise summaries of deeply revealing news articles on corporate corruption from reliable major media sources.
OpenAI this week quietly deleted language expressly prohibiting the use of its technology for military purposes. Up until January 10, OpenAI’s “usage policies” page included a ban on “activity that has high risk of physical harm, including,” specifically, “weapons development” and “military and warfare.” That plainly worded prohibition against military applications would seemingly rule out any official, and extremely lucrative, use by the Department of Defense or any other state military. The new policy retains an injunction not to “use our service to harm yourself or others” and gives “develop or use weapons” as an example, but the blanket ban on “military and warfare” use has vanished. OpenAI spokesperson Niko ... Felix [said] that OpenAI wanted to pursue certain “national security use cases that align with our mission,” citing a plan to create “cybersecurity tools” with DARPA, and that “the goal with our policy update is to provide clarity and the ability to have these discussions.” The real-world consequences of the policy are unclear. Last year, The Intercept reported that OpenAI was unwilling to say whether it would enforce its own clear “military and warfare” ban in the face of increasing interest from the Pentagon and U.S. intelligence community. “Given the use of AI systems in the targeting of civilians in Gaza, it’s a notable moment to make the decision to remove the words ‘military and warfare’ from OpenAI’s permissible use policy,” said [former AI policy analyst] Sarah Myers West.
Note: Learn more about emerging warfare technology in our comprehensive Military-Intelligence Corruption Information Center. For more, see concise summaries of deeply revealing news articles on corporate corruption from reliable major media sources.
Leading up to the August Republican presidential primary debate ... An RNC official told Google via email that the debate would be streaming exclusively on the upstart video platform Rumble. The August 23 debate was broadcast on Fox News and streamed on Fox Nation, which requires a subscription, while Rumble was the only one to stream it for free. On the day of and during the debate, however, potential viewers who searched Google for “GOP debate stream” were returned links to YouTube, Fox News, and news articles about the debate, according to screen recordings. Rumble was nowhere on the first page. For Rumble, which is currently in discovery in an antitrust lawsuit against Google in California, this is a case of Google suppressing its competitors in favor of its own product, YouTube. YouTube is owned by Google, and it has regularly been the subject of anticompetitive allegations from rivals, who charge that Google unfairly and illegally favors YouTube in its search algorithm. Google, in fact, is in the middle of a landmark antitrust trial, charged with anticompetitive practices by the Department of Justice. The company would not have been required by antitrust law to promote [Rumble's] link. It would, however, be barred from suppressing the competitor’s link from organic results. The fact that Rumble’s link did not appear on the first page even though it was the most relevant link the search could return means either the search engine failed at its task or the link was suppressed.
Note: For more along these lines, see concise summaries of deeply revealing news articles on corporate corruption and media manipulation from reliable sources.
The EARN IT Act [is] a bill designed to confront the explosion of child sexual abuse material (CSAM) online. EARN IT would help address what is, disturbingly, a common experience for young users: routine exposure to predatory targeting, grooming, sexual violence, prostitution/sex trafficking, hardcore pornography and more. A New York Times investigation revealed that 70 million CSAM images were reported to the National Center for Missing and Exploited Children (NCMEC) in 2019–up from 600,000 in 2008–an "almost unfathomable" increase in criminality. The EARN IT Act restores privacy to victims of child sexual abuse material and allows them to sueâ€those who cause them harm online, under federal civil law and state criminal and civil law. It also creates a new commission to issue guidelines to limit sex trafficking, grooming and sexual exploitationâ€online. CSAM still exists because tech platforms have no incentive to prevent or eliminate it, because Section 230 of the Communications Decency Act (passed in 1996, before social media existed) gives them near-blanket immunity from liability. While some in the technology sector [are] claiming EARN IT is a threat to encryption and user privacy, the reality is that encryption can coexist with better business practices for online child safety. We can increase security and privacy while refraining from a privacy-absolutism that unintentionally allows sexual predators to run rampant online.
Note: To understand the scope of child sex abuse worldwide, learn about other major cover-ups in revealing news articles on sexual abuse scandals from reliable major media sources.
Ties between Silicon Valley and the Pentagon are deeper than previously known, according to thousands of previously unreported subcontracts published Wednesday. The subcontracts were obtained through open records requests by accountability nonprofit Tech Inquiry. They show that tech giants including Google, Amazon, and Microsoft have secured more than 5,000 agreements with agencies including the Department of Defense, Immigrations and Customs Enforcement, the Drug Enforcement Agency, and the FBI. Tech workers in recent years have pressured their employers to drop contracts with law enforcement and the military. Google workers revolted in 2018 after Gizmodo revealed that Google was building artificial intelligence for drone targeting through a subcontract with the Pentagon after some employees quit in protest, Google agreed not to renew the contract. Employees at Amazon and Microsoft have petitioned both companies to drop their contracts with ICE and the military. Neither company has. The newly-surfaced subcontracts ... show that the companies' connections to the Pentagon run deeper than many employees were previously aware. Tech Inquiry's research was led by Jack Poulson, a former Google researcher. "Often the high-level contract description between tech companies and the military looks very vanilla," Poulson [said]. "But only when you look at the details ... do you see the workings of how the customization from a tech company would actually be involved."
Note: For more along these lines, see concise summaries of deeply revealing news articles on corruption in government and in the corporate world from reliable major media sources.
Frances Haugen spent 15 years working for some of the largest social media companies in the world including Google, Pinterest, and until May, Facebook. Haugen quit Facebook on her own accord and left with thousands of pages of internal research and communications that she shared with the Securities and Exchange Commission. 60 Minutes obtained the documents from a Congressional source. On Sunday, in her first interview, Haugen told 60 Minutes correspondent Scott Pelley about what she called "systemic" problems with the platform's ranking algorithm that led to the amplification of "angry content" and divisiveness. Evidence of that, she said, is in the company's own internal research. Haugen said Facebook changed its algorithm in 2018 to promote "what it calls meaningful social interactions" through "engagement-based rankings." She explained that content that gets engaged with – such as reactions, comments, and shares – gets wider distribution. "Political parties have been quoted, in Facebook's own research, saying, we know you changed how you pick out the content that goes in the home feed," said Haugen. "And now if we don't publish angry, hateful, polarizing, divisive content, crickets." "We have no independent transparency mechanisms," Haugen [said]. "Facebook ... picks metrics that are in its own benefit. And the consequence is they can say we get 94% of hate speech and then their internal documents say we get 3% to 5% of hate speech. We can't govern that."
Note: For more along these lines, see concise summaries of deeply revealing news articles on media manipulation from reliable sources.
Defense industry tycoon Palmer Luckey [is] the 33-year-old co-founder of sprawling weapons manufacturer Anduril Industries. A longtime GOP donor and activist who spoke on condition he not be identified in order to preserve his political relationships said Luckey was now the “godfather of Republican fundraising on the west coast”. “If you’re raising campaign money in California, you have to go to Palmer Luckey,” he said. “Luckey is the top tier.” Anduril, launched just nine years ago, has emerged as a major military contractor for the Trump administration, working on contracts for everything from drones to autonomous jet fighters to space-based interceptors. In March, the US army awarded a $20bn sole-source contract to Anduril for a variety of drones it sells, though it is not clear which of its weapons systems are operational or have been deployed for combat in US operations. Anduril, though it is not widely known as a space contractor, announced in May it would work, with a team of other companies, on what may be the mother of all contracting opportunities offered by the Trump administration, the so-called “Golden Dome” anti-missile space-based interceptor effort. Estimates for the Golden Dome range between $185bn and $1.2tn. Anduril says in Trump’s first year in office it doubled its revenue – a growth rate of 100%, but still only reached $2.2bn for 2025, which is barely a blip in the Pentagon’s trillion-dollar budget.
Note: For more along these lines, read our concise summaries of news articles on Big Tech and warfare technology.
Hundreds of contractors working on a project for Meta were instructed to pose as minors online and probe how competitor chatbots responded to prompts involving suicide, sex, eating disorders, and other high-risk subjects, according to internal documents and five people familiar with the project. The prompts were often designed to push the chatbots toward responses their safety systems were supposed to refuse. A single round of testing completed in August 2025 saw more than 45,000 prompts run through the rival chatbots. The companies behind the chatbots weren’t aware of the testing. Testing competitors’ products is not, by itself, unusual in the artificial intelligence industry. Former contractors who worked on the project described several aspects as alarming. According to one former worker, employees feared the possibility they could be generating or preserving child sexual abuse material if a chatbot responded to certain sexual prompts involving minors. Another says they worried the project amounted to secretly taking material from competitors’ systems to potentially feed back into Meta’s system. While a dataset of thousands of youth-safety prompts could be useful for comparing how often chatbots refuse harmful requests, the scale and opacity of [this project, known internally as] Cannes, along with the lack of disclosure to the companies being tested, made it very different from other public safety benchmarks.
Note: For more along these lines, read our concise summaries of news articles on Big Tech and mental health.
Humanity has officially passed a grim milestone: the first confirmed death at the hands of a fully-autonomous robot weapon. A senior figure involved in the Ukrainian defence industry admitted that a fully autonomous drone deployed to the front lines of the Russo-Ukranian war has claimed human lives. The event has been kept under wraps for two years, Ukrainian drone maker Alexander Kokhanovskyy said. The device itself was a quadcopter drone designed to zip to the front line, loiter around a perimeter for roughly 10 minutes, then launch into “terminator mode.” This mode, the Ukranian drone expert explained, involves the use of an unidentified AI model to automatically seek out and attack human targets. During the operation in question ... it identified and ultimately attacked “a couple of [Russian] soldiers” and “one truck.” “We just launch it and we know everything will be dead,” Kokhanovskyy [said]. “There is no connection to the drone at all ... everything it sees will be killed.” While this marks the first known strike of its kind, it’s fair to say the world’s militaries have been inching toward this for a while. The Israeli Defence Forces, for example, have long deployed loitering suicide drones to hover over various demarcation zones across Palestine. More recently, US military operatives used Anthropic’s Claude to select bombing targets across Iran. If Kokhanovskyy is to be believed, humanity has crossed a threshold that can never be uncrossed.
Note: This is being called AI's "Oppenheimer moment". For more along these lines, read our concise summaries of news articles on war and artificial intelligence.
Jeff Bezos’ Washington Post is facing a scathing new class-action lawsuit accusing it of using surveillance pricing to gouge loyal readers – in a case that attorneys believe could rack up millions in damages. Since the mid-2010s, WaPo has “covertly harvested” subscriber data, using “deeply personal information” to determine how much they could squeeze out of each loyal reader, according to the suit, which was filed Thursday in the Superior Court of the District of Columbia. Readers largely expected their personal data would be used for “mutually beneficial purposes” like “relevant advertisements” – and did not consent to their personal information being used to hike prices on their subscriptions, the suit alleged. After billionaire Bezos bought WaPo in 2013 for $250 million, the paper started heavily investing in technology and digital subscriptions, the suit noted. “The more loyal a reader became, the more data The Post could gather to estimate how much more that person might tolerate paying at renewal,” the complaint said. “Rather than rewarding loyalty, The Post’s system converted subscribers’ engagement into leverage against them.” The suit also alleged WaPo might be collecting extra information from subscribers’ use of “affiliates” – including Amazon, Bezos’ e-commerce giant. Last year, the publication was forced to reveal it was engaging in surveillance pricing techniques because of a 2025 New York disclosure law.
Note: Read more about the rise of AI surveillance pricing. For more along these lines, read our concise summaries of news articles on Big Tech and mass media.
AI is projected to generate nearly unfathomable amounts of revenue. Any mention of AI tends to be accompanied by warnings that deeper jobs cuts across many more industries are coming for us all. Jensen Huang, CEO of chip giant Nvidia, said in 2025: “Every job will be affected, and immediately. It is unquestionable. You’re not going to lose your job to an AI, but you’re going to lose your job to someone who uses AI.” Increasingly, people young and old flock to a new gold rush in Silicon Valley to toil away on AI-fueled startups. If AI’s worst-case scenario for tech jobs plays out ... that’s still nowhere near the apocalyptic future of labor that many fear. “Is it, in fact, going to destroy all of the jobs?” Naidu asked. “I’m not convinced. Even take software. Software is only about 4 to 6% of GDP. So it’s a lot, but it’s not like the whole economy can be replaced by Claude Code.” Convincing people that AI will replace human workers in droves is a clever marketing tactic. Not only does it stoke rabid investor speculation, but it distracts from a more realistic application of AI: to surveil and micromanage employees to squeeze yet more productivity out of them, all the while pressuring them to feel grateful that they have any kind of work. Gig workers, the people who pick you up in Ubers and deliver your food on platforms like DoorDash, have already been the guinea pigs for this kind of algorithmic management, and labor experts predict it will spread.
Note: For more along these lines, read our concise summaries of news articles on AI and media manipulation.
For decades, Congress has tried and failed to give Americans control over their own personal data: the right to see it, correct it, and delete it at will. This inaction has left Americans with no recourse against misuse of their own data, while the data broker industry quietly continues to collect and sell the personal information of millions, operating in a largely unchecked gray market. Now, two new bills, the SECURE Data Act and the GUARD Financial Data Act, offer the latest test of whether Washington can step up and finally pull data brokers out of the shadows and into the reach of the law. Efforts to prevent the SECURE Data Act — or any federal protections — from being enacted are currently on full display. Exacerbating the situation and further endangering consumers, there is an entire category of companies that have deliberately avoided being classified as data brokers in an effort to skirt even the patchwork of state-level regulations. Unlike traditional data brokers, massive data aggregators don’t sell your name and address to the highest bidder. Instead, they operate quietly, harvesting your data from across the internet, then assembling it into risk scores, behavioral profiles, and assessments of your creditworthiness. These opaque calculations increasingly govern your real-world outcomes, including whether you’re approved for a mortgage, the interest rates on your auto loan, and what services or products are marketed to you.
Note: For more along these lines, read our concise summaries of news articles on Big Tech and the disappearance of privacy.
Former Meta employee Sarah Wynn-Williams [was] silenced by Meta’s legal threats to bankrupt her if she spoke. Wynn-Williams has written a book, Careless People, about her time at Meta (then Facebook), where she was an early director of global public policy. But Meta does not like the book. It has done everything in its power to stop it, including seeking an emergency arbitration order that prevents Wynn-Williams from promoting the book, and threatening punitive damages. These serve both to punish Wynn-Williams for writing it, and to send a warning to any future critic. A certain kind of libertarian responds by saying that Meta is not “censoring” Wynn-Williams, because only governments can censor. A certain kind of lawyer may say she brought this on herself by signing a contract agreeing not to criticise Meta. Private censorship is real and, in the time we live in, often more impactful than the public kind. Not all contractual provisions are, or should be, enforceable. You cannot write an enforceable contract to sell a child, to bind someone never to marry or to give up other fundamental rights. Why should the right to speak critically be any different? A contract in which someone agrees never to criticise their employer should be void and unenforceable. That is why we need legislation that makes clear a simple principle: that the free-speech right to criticise your employer is important, fundamental and cannot be sold.
Note: Meta whistleblower Sarah Wynn-Williams once told US senators that the company targeted teenage girls with beauty and weight-loss advertisements during moments of heightened vulnerability such as after deleting a selfie. According to her testimony, Meta could detect when users were feeling "worthless," "helpless," or like a "failure," and then make that information available to advertisers. For more along these lines, read about a new nonprofit called Psst, which is designed to make it safer for Big Tech whistleblowers to report wrongdoing without immediately exposing themselves to retaliation.
The former chief investigations counsel for the House Oversight Committee has been helping to prepare Bill Gates, the billionaire co-founder of Microsoft, to testify privately in the panel’s Jeffrey Epstein investigation on Wednesday. Representative James R. Comer of Kentucky, the Republican chairman of the committee, formally requested in March that Mr. Gates appear before the committee for a transcribed interview. His request came after files released by the Justice Department showed that Mr. Gates met with Mr. Epstein, the convicted sex offender, multiple times and that his closest advisers were in frequent contact with the disgraced financier until 2019, the year of his death in prison. In preparing for the deposition, Mr. Gates has turned to Jake Greenberg, who until December was spearheading the oversight panel’s Epstein inquiry in his role as the committee’s top investigative official. Mr. Gates’s close relationship with Mr. Epstein has roiled his foundation, which has authorized an outside review of its ties to Mr. Epstein. Representative Suhas Subramanyam, Democrat of Virginia, said in an interview that he wanted to know what Mr. Gates “knew of Epstein’s crimes, and the nature and extent of their relationship.” He added, “Epstein was known for befriending and even blackmailing rich and powerful men, and I want to know if Gates was one of them.” Mr. Gates has sought out powerful inside players to help him weather the scrutiny. He hired John Moran, a former lawyer for the Justice Department, who helped him secure an agreement with the committee for him to appear off camera, The Wall Street Journal reported.
Note: Don't miss part one and part two of our investigations into the Epstein files so far. For more along these lines, read our concise summaries of news articles on Big Tech and Jeffrey Epstein.
Some of the largest data-collecting companies in the United States—including major AI vendors, data brokers, defense contractors, and dating apps—rely on deceptive methods to keep consumers from opting out of the sale and sharing of their personal information. Researchers at [the Electronic Privacy Information Center] audited the opt-out processes of 38 major data companies and documented at least eight distinct categories of manipulative design: Opt-out forms that don't actually let users opt out of the sale of their data. Links that are buried in fine print and missing from homepages. Consumers routed through multiple separate forms to complete a single request. And requirements that users create accounts or pay for subscriptions before opting out at all, among others. Major companies offering large language models, such as Google, Meta, and OpenAI, fail to clearly link their opt-out forms from their homepages or privacy policies, according to the report, and several require consumers to submit multiple separate forms to complete a single request. OpenAI's form, when a consumer finds it, does not offer a way to opt out of the sale or transfer of personal data. What it offers instead is an option to “remove personal information from ChatGPT responses,” which EPIC says is a filter on the chatbot's output, not the removal of any underlying data. Researchers found that the people-search brokers they audited—Spokeo, Whitepages, and National Public Data—do not offer consumers a way to opt out of the sale or transfer of their data at all. Instead, the companies offer a process for removing individual listings by URL, one at a time, with no commitment to stop selling that same person's information in the future.
Note: The owner of a data broker company once bragged about having highly detailed personal information on nearly all internet users. For more along these lines, read our concise summaries of news articles on Big Tech and the disappearance of privacy.
A small group of experts [were] enlisted by A.I. companies to vet their products for catastrophic risks. Some have shared with The Times more than a dozen chatbot conversations revealing that even publicly available models can do more than disseminate dangerous information. The virtual assistants have described in lucid, bullet-pointed detail how to buy raw genetic material, turn it into deadly weapons and deploy them in public spaces, the transcripts show. The U.S. government has long planned for powerful adversaries unleashing deadly bacteria, viruses or toxins in the American population. Since 1970, there have been a few dozen, fairly small biological attacks around the world, such as the anthrax-laced letters that killed five Americans in 2001. Despite perennial warnings, a major catastrophe has not happened and remains unlikely, most experts say. But even if the probability is low, an effective biological weapon could have an enormous impact. Dozens of experts told The Times that A.I. is one of several recent technological advances that have meaningfully increased that risk by expanding the pool of people who could cause harm. Still, A.I. users would need some real-world expertise to follow a bot’s instructions. Some research, including a study backed by A.I. companies, has found that while chatbots can help novices learn certain lab skills, the technology isn’t particularly helpful for carrying out the range of complex tasks needed to make a virus from scratch.
Note: For more along these lines, read our concise summaries of news articles on AI, biotech dangers, and science corruption.
Important Note: Explore our full index to revealing excerpts of key major media news articles on several dozen engaging topics. And don't miss amazing excerpts from 20 of the most revealing news articles ever published.
