{
  "type": "article",
  "title": "How a Hardware Theft Suit Slammed Into OpenAI's Worst Week Yet, While a Flesh-Twisting Parasite Swept Across 30 States",
  "summary": "In a single week, Apple hit OpenAI with a lawsuit over stolen hardware secrets, OpenAI's own staff bankrolled a rival super PAC, New York passed the country's first data-center moratorium, and America logged its worst-ever cyclosporiasis outbreak.",
  "content": "Some weeks in the tech world seem to crack open all at once, and this was exactly that kind of week for OpenAI. On one side, Apple slapped the company with a serious lawsuit accusing it of stealing confidential hardware secrets. On the other, a group of OpenAI's own employees quietly opened a front against their own boss by launching a super PAC. At the same time, New York became the first state in the country to hit the brakes on building large data centers, fresh questions surfaced about how DOGE used AI, and a parasitic infection tore across the United States, triggering an outbreak of weeks-long, explosive diarrhea. Let's walk through each of these stories one by one, and what they actually mean for ordinary people.\n\nStart with OpenAI, because the company dominated the headlines all week, and not for particularly happy reasons. Last Friday, Apple filed a lawsuit against OpenAI. The accusation is that OpenAI stole highly sensitive information, including unreleased iPhone parts and prototypes, confidential designs, and documents about secret projects. For any company, that charge is heavy on its own. But it gets much messier because Apple says this alleged theft happened mostly through people who used to work at Apple.\n\nTang Tan and the more than 400 former Apple employees\nThe lawsuit names OpenAI's chief hardware officer, Tang Tan, directly. Tan spent 24 years of his career at Apple. Apple alleges that Tan encouraged people who were leaving the company to bring proprietary information and unreleased technology with them. According to the lawsuit, OpenAI has hired more than 400 former Apple employees. That figure alone shows just how high the stakes are in this hardware fight.\n\nThis is part of an Apple playbook that we haven't seen used in a while, though it is something Apple has done or threatened to do before. A telling example is Tony Fadell, who was a longtime Apple guy before he went on to start Nest, the thermostat company. As the story goes, when Fadell started Nest and hired several hundred Apple employees, Steve Jobs called him, threatened to sue, and screamed bloody murder. What Fadell said about it fits this whole situation perfectly. He said, \"It's my job to hire great people. It's your job to keep them.\"\n\nIt's worth remembering that Apple is a fairly litigious company. It has famously sued employees for supposedly leaking things or taking proprietary material with them when they left. Companies do this occasionally, but lawsuits over leaks in particular are rare. Still, this is the area Apple cares most about. It is so secretive about its products that the moment that material starts slipping out of its hands, it moves quickly to clamp down.\n\nThe real goal isn't damages, it's slowing OpenAI's hardware push\nLook closely and this lawsuit doesn't seem to be about squeezing money out of OpenAI. What Apple really wants is to slow down OpenAI's hardware ambitions. Apple keeps going all in on the iPhone as the primary computing platform for the AI era. So if an audio-first platform emerges that works better for things where you don't need to look at a screen and can just talk to an agent, that could be genuinely intimidating for them.\n\nThe device itself is expected to look like a speaker. It's said to have some motorized elements so it can move around in some capacity, and the design has been compared to a Furby, which is probably about right. But the question is whether OpenAI can really pull it off if it's banking on hardware, and there's plenty of skepticism there. Ultimately, Apple can still make a speaker. And because Apple is a more neutral player in all this, it could make a speaker that lets you choose from a few different AI options, or one that runs Siri's AI. In the end, it comes down to the fact that the device you use most of the time is still going to be the phone.\n\nThat's been true for a long time. Apple itself has tried to get away from this very screen-heavy life we're all living, with its face computer, the Vision Pro. But the Vision Pro arrived and faded from the conversation just as fast. The same problem has dogged other AI hardware devices, like the famous pin from Humane AI. The truth is that plenty of things are simply better to do on a screen, and that isn't going to change. That said, if agents and voice mode ever work really, really well, there are some things you could just ask an agent to execute for you, where you'd prefer not to stare at a screen the whole time. For anyone with a tortured relationship with their screens, something like that could be a relief. But it's very hard to execute correctly, and so far no product has managed it.\n\nTo ground all of this, OpenAI last year paid $6.5 billion to acquire a startup called IO Products, which was co-founded by longtime Apple executives, including Tan, Scott Cannon, Evans Hankey, and, most famously, Jony Ive. It's a huge investment in this space, and losing all of these people obviously stings for Apple. Apple has already lost AI researchers to other companies, and now it's hemorrhaging hardware talent too, which hits it right where it hurts.\n\nThe most tantalizing part of this whole process hasn't even arrived yet. A lawsuit means discovery, and discovery means getting to read piles of emails in which these companies talk trash about one another and about themselves. When lawyers square off over an intellectual property dispute, that's usually where the real drama spills out.\n\nOpenAI's second headache: a revolt from inside the house\nOpenAI's troubles weren't limited to the Apple lawsuit. This week it emerged that some OpenAI staffers are funding a rival super PAC to push for tighter regulation of frontier AI labs. It's a counterweight to all the money that people like OpenAI executive Greg Brockman and others have poured into boosting AI and tearing down guardrails. This new super PAC is called Guardrails Alliance, and it launched last month with $5 million in initial funding.\n\nThe super PAC bills itself as a populist effort by tech workers, labor unions, and other groups. Its key purpose is to serve as a counterweight to the $100 million Leading the Future fund, which lobbies against guardrails. Even if $5 million looks small next to that $100 million, the mere fact that it exists is significant.\n\nThis moment has felt like it was coming for a while, ever since it became clear that Greg Brockman was donating so much money to MAGA and Trump-related causes. It's really all about boosting AI and AI policy. Brockman's view, shared by many others in the industry, is that Trump and those aligned with him are more favorable toward a growth-at-all-costs, AI-first policy agenda. But the reality is that these companies are overwhelmingly filled with younger employees who lean pretty liberal. That tension has the potential to get genuinely messy.\n\nOpenAI still has a culture where people are willing to push back and speak out. Its internal Slack has that earlier-tech feel of people really working out problems in public, even though this stuff tends to leak so much. At the same time, the makeup of the company is changing fast. Over the past year or two, a wave of newer employees has come in, and there's a sense that OpenAI is really shifting. There is now a Trump-aligned contingent inside the company and a more liberal contingent on the other side.\n\nGrassroots versus billionaire, all inside one company\nIn many ways this is a grassroots-versus-billionaire fight, except it's playing out entirely within a single company. We sometimes talk about Silicon Valley as a monolith, when it is clearly nothing of the sort. There are all kinds of people at every rung, with very different views. One striking detail is that one of the largest donors was a research engineer, Juan Felipe Cerón Uribe, who gave $200K. He has spent years working on the company's strategies for mitigating the potential societal harms caused by AI.\n\nThat raises a real question about whether companies like this have any stopgaps for situations like this. OpenAI is a relatively mature company. It has processes and an HR department. That doesn't mean this kind of thing can't happen. It absolutely can, and we've seen it happen before. But until it becomes a major issue, until this fight actually starts to impact the work and OpenAI's ability to ship new models, the company will probably stick to the line that \"we have an open culture, people are allowed to disagree, this is totally fine.\" And it will try to spin it almost as a positive.\n\nGuardrails Alliance versus Leading the Future: where the money should go\nIn the grand scheme of things, $5 million is not a lot of money. The super PAC has a larger goal of raising $15 million during this election cycle. So the question is what it can reasonably accomplish with that. It's worth pointing out that Anthropic employees have also started a separate PAC trying to combat Leading the Future, called Public First Action, which has something like $20 million. So there is other funding and there are other people in the AI space making moves against the Greg Brockmans of the world. But when you add it all up, it's roughly $25 million versus $100 million, which is still an enormous gap.\n\nThat gap suggests these employees need to be working on down-ballot, local campaigns. You can get a lot of bang for your buck in smaller states with less spending. Right now the races everyone wants to talk about are Texas and Maine, which require very expensive ad buys. But take a state like Maine, where there's a lot of local volunteer power that would be very interested in getting involved for very cheap, along with local organizations that can mobilize people just as cheaply. Radio spots in Maine versus radio spots in Austin are worlds apart in cost.\n\nUltimately, the outcome will come down to who is advising them. If these employees are being told to jump into the biggest, most expensive races in places like New York, that would be a mistake. By all means make your thoughts known on AI regulation, New York data centers, and all of these issues. But if you want to create bigger moments and movements, you have to go down-ballot. The other real question is whether this is a play for notoriety, a way to say, \"Look, there are plenty of OpenAI employees who don't agree with you,\" or a genuine attempt to affect races and perhaps turn the Senate blue. And in some ways, regardless of the outcome or the money behind it, it can be just as important to be loud and visible in saying, \"Hey, we don't all think the same about this.\"\n\nDOGE, housing policy, and AI's black box\nSpeaking of AI in politics, DOGE, the Department of Government Efficiency started by Elon Musk, is back in the news, and for a very specific reason. Members of DOGE working at the Department of Housing and Urban Development, HUD, used AI to inform policy decisions. That was uncovered last year, but now the agency appears to be denying Freedom of Information Act requests for information on the development and use of these AI tools, and how they actually shaped those policy decisions.\n\nThis all came out through documents obtained via a FOIA request from Democracy Forward, a nonprofit legal organization. And it's pretty concerning. You'd assume that a year after some of the dust settled, we'd be able to get a little more information. But from one government agency to the next, there has been a steady stream of stonewalling on further details about what DOGE was up to, the access it had, and the materials it used. It's very much about the fact that they used AI, but also how they used AI, what is still left in the systems, what regulations were there, what got blown past, and who knew about this. So many important details are simply missing, and that's especially frustrating now that people are going back and trying to untangle the tangled web that was left behind. It's frustrating for employees, understandably, and it should be frustrating for anyone who will be affected by housing policy in the United States over the next 20 years.\n\nThis speaks to an issue we're all grappling with. AI tools can be used in ways that are entirely legitimate and helpful. If you were told they were using AI for research and to check their sources, that would seem smart. But they can also be used in ways that possibly introduce discrimination, or in ways some would call cheating. The sheer range of tools and how they're used raises a lot of questions, questions the public is entitled to some form of answers on. But right now, those answers are being withheld.\n\nIn this specific case, HUD employees found that these DOGE staffers were using AI to identify agency rules for potential rescission or contract cancellations. And this was happening across government. Broadly, it was like saying, \"Here's this massive list of contracts, what do you think we should cut here,\" and then just seeing what happens. There aren't many details on everything that got cut, the way it got cut, or what we're missing now as a result.\n\nThe people doing this matter too. Christopher Sweet, who was then a third-year student at the University of Chicago, was one of the people really involved in this effort. So was Scott Langmack, who came to DOGE from a property-technology startup called Kukun. So these weren't necessarily people who were especially well versed in government work. And on top of that, the idea of running through massive government contract lists with AI and going, \"Yep, good to go,\" is striking in itself.\n\nThe \"AI exemption\" that doesn't actually exist\nIn this case, there's a claim to an AI exemption. The argument is essentially, \"We can't possibly release these to you because it was AI, and AI is exempt from FOIA.\" But that isn't the case. There is no AI exemption under FOIA. AI talking to people and AI talking to each other is not privileged communication, which is the argument that was being made. There are no laws in the US that require the government to disclose whether AI was used in the creation of rules, policies, or regulations. And so a FOIA request like this is the only way you'd ever find out. Claiming an AI exemption simply turns the whole thing into a black box you can't see into.\n\nThe question isn't just how the tools were used, but how the tools made the decisions they made. Anyone who has used AI agents at all knows you can't just say, \"Identify regulations we should potentially take X, Y, Z action on.\" You have to give them a lot of context. And even then, it's very important to interrogate the tools and ask, \"How did you make that decision,\" so you can understand what the AI is taking into account and what it's ignoring. These tools are not completely neutral. They're built by humans, and they have biases baked in. One would sincerely hope, without much hope, that DOGE was looking at all of that and not just unleashing AI on these systems willy-nilly.\n\nA crucial part of this story is that Scott Langmack, from the property-technology startup Kukun, is now the executive director of deregulation AI at the Office of Management and Budget, which falls under the executive office of the president, according to his LinkedIn. So it's not as if this person has vanished into obscurity. His decision-making continues inside that same black box, because they won't tell us anything about it. It very much felt like DOGE swept in and swept out in the same wave of the year, but that's really not the case. The people who got deeply involved and entrenched these policies from the beginning are still not just in government, but officially in government, in serious roles with real pay scales.\n\nNew York's historic data-center moratorium\nNow one more AI story. New York became the first state in the country to impose a statewide moratorium on building new data centers. The moratorium was signed as an executive order by New York Governor Kathy Hochul, and it puts a one-year pause on the development of large-scale data centers while the state develops new environmental and energy-grid standards.\n\nHochul's own words capture the thinking behind it. \"The bottom line is progress shouldn't arrive with a higher utility bill, depleted water supply, or noise pollution. So we have no choice but to address these challenges created by these massive facilities. That is why today I'll be signing the nation's first ever statewide moratorium on hyper-scale data centers.\"\n\nThere are a few important caveats. First, the moratorium applies only to large-scale data centers, meaning 50 megawatts or more. Second, it doesn't affect facilities that have already started construction or already hold permits. And third, it will lift once the state finalizes a generic environmental impact statement, which could be a year or more out.\n\nThis moratorium is really wrestling with two stories. There's the national story, which is that this country genuinely needs AI. It's a big part of the economy, and in many ways it's the big bet we've taken. It feels like there isn't a lot else right now, especially with the electric-vehicle push having largely fizzled. And then there's the local story, which is that it is deeply uncomfortable, and frankly miserable, to live near a data center. It spikes your energy bill and it sucks up a lot of water. Those two things are going to keep colliding.\n\nLarge-scale data centers are deeply unpopular\nRight now, large-scale data centers are just deeply unpopular. It feels like every other headline is about one community rebelling against a project, and then another community rebelling against another. Other states have tried and failed to pass a similar moratorium, so it will be interesting to see how New York's plays out.\n\nPresident Trump weighed in on the decision. At 12:39, he posted on Truth Social, \"One of the biggest driving forces in the future for jobs are data centers. They are big, strong, bold and money machines for the state in which they are built. Governor Kathy Hochul, for political reasons, has terminated all data centers being built or to be built in New York State. These companies are now being sought in Alabama, Florida, Texas, Arizona, and many other states, but the taxes and the job amount to liquid gold. New York State has made a terrible decision.\" He went on, \"The radical left 'dumbocrats' must not be allowed to cause us to lose data centers, AI and all of this incredible new technology to China and other countries.\"\n\nFrom a political angle, whoever told Donald Trump that data centers would be his winning issue for the midterms is a curiosity in itself. There actually is an argument he could have made that fits squarely in his wheelhouse, an America-first argument. It's that AI is the next space race, that we need to beat China, and that we need data centers to do it. Trump does eventually get there when he mentions China, but the whole post is littered with errors, and it misses a key part of his electorate, one that isn't only in blue states, that's asking how on earth it's supposed to pay for this and live with it.\n\nTwo very different lenses on data centers\nThere are two clear ways to look at this. One view is that the focus should be on making data centers safer, better, and less harmful to the communities near them. They should be built as far from where people live as possible, and made as energy-efficient as possible. By this logic, a blanket moratorium, which again isn't exactly what New York did, would be a mistake, simply because of how dependent the entire economy currently is on AI and how important these data centers remain for the AI race and boom.\n\nThe other view disagrees and argues that the carve-outs here make sense. You can still build anything up to 50 megawatts, meaning you can still build some pretty sizable mega centers. This view is grounded in lived experience too, because someone can find a big data center going up just 10 miles from their house, right next to where an animal shelter was going to move. These projects often get decided before anyone realizes it's even happening. That's the reality in a lot of places in this country, where the deals were made a while ago, or when they do happen, the community doesn't find out until the permitting is done that this is even coming. So some sort of pause, some sort of moratorium, some sort of restrictions make sense, at least until we figure out a way to inform people about what's going on and settle on a better plan for where these things go and the impact they'll have.\n\nTalking to a lot of industry people, it's genuinely surprising how constrained startups feel by compute. Places like SpaceX and OpenAI have a lot of access to data centers and computers, while everyone else is fighting tooth and nail for access and needs far more than they can currently get. That makes the AI-bubble talk feel a little less alarming, because at least in Silicon Valley the appetite still feels enormous.\n\nThis is shaping up to be a real issue in the midterms, at least in some places. And what's interesting is that, despite Trump's post, it doesn't seem cleanly divided along party lines. Plenty of Republicans and Democrats have both been opposing data centers in their communities, asking for reform, and asking for moratoriums. That's because community impact is community impact, whether it's a blue state or a red one.\n\nCyclosporiasis: America's worst outbreak on record\nNow a story that isn't for the faint of stomach, but is an important public-health story all the same. The US is in the middle of its worst outbreak on record of a nasty parasitic infection called cyclosporiasis, which causes explosive, weeks-long diarrhea. It's spreading fast, with cases now confirmed in more than 30 states.\n\nCyclospora is a microscopic parasite that causes a pretty bad intestinal illness, and the illness itself is called cyclosporiasis. People get it by consuming food or water contaminated with the parasite. The CDC suspects there may already be close to 7,000 cases nationwide. Michigan makes up the bulk of those, and on Wednesday its case count was updated to more than 3,700 cases. For context, in a typical year states might report a handful of cases to a few dozen. So this is really unusual. The actual case count is almost certainly higher, because most people don't seek medical care when they get diarrhea, and even when they do, labs don't routinely test for this particular parasite. But now that there's awareness of it, if you have symptoms, you should definitely ask your doctor to test specifically for this.\n\nIt's important to understand this is not an ordinary bout of diarrhea. This isn't just a couple of bad bowel movements. One man's case got so severe he was on the toilet 40 times in a single day. This is far beyond what a casual bout would do to a person.\n\nThis is not a one-and-done type of thing. It can last days, possibly up to a week or two of diarrhea. One of the big concerns with this prolonged diarrhea is dehydration. So people should be aware of that if they think they have symptoms, and keep hydrated. If you are older, if you are immunocompromised, or if you are particularly at risk for severe dehydration, you should really be concerned about this and be on the lookout for dehydration.\n\nThe cause is still a mystery, but suspicion falls on lettuce\nThe exact source of the outbreak is still a mystery. The CDC is still investigating several food products, although Michigan health officials have said that lettuce keeps coming up as a common product during their investigation. In general, it's well known that lettuce and leafy greens are a common source of other foodborne illnesses. Romaine lettuce, for instance, has always been linked to E. coli outbreaks and salmonella, while other leafy greens are linked to norovirus. So it wouldn't be a surprise if lettuce ends up being the culprit here.\n\nOne reason lettuce is so problematic is that it's eaten raw. You don't hear about disease outbreaks from Brussels sprouts because we cook them. Another reason is that lettuce and other leafy greens have small folds and crevices that make it difficult to wash pathogens off completely.\n\nHow to protect yourself\nSo how do you get rid of it? This is where it gets tricky, because the cyclospora parasite is resistant to bleach and chlorine, and vinegar doesn't kill it either. One thing you should know is that if you're buying bagged lettuce that says it's pre-washed, wash it anyway. This parasite is very good at latching onto the surfaces of fruits and vegetables. So right now you really need to get in there and give your greens a good scrub. You should be scrubbing down all of your raw vegetables, but especially lettuce.\n\nIn terms of practical guidance, if you're buying lettuce, scrub it clean. When it comes to eating at home versus out, maybe don't eat salads out, because you don't know what the cleaning process was like. If you trust yourself to clean it properly, then by all means still go for the salad at home if you like salad. Whether you want to risk it is up to you.\n\nThe role of government cuts and a weakened CDC\nYou can't really talk about this without talking about government cuts in the healthcare space. Last year, cutbacks to the CDC's FoodNet came to light. HHS has pushed back a bit on the idea that these kinds of funding cuts are to blame, saying this monitoring has continued throughout their systems. So to what extent do the cuts actually play a role?\n\nA former CDC official has said that FoodNet wasn't designed for this kind of real-time outbreak detection or response. Its function is more about tracking longer-term trends. But the bigger issue here is the huge staffing cuts to the CDC. It was reported last year that about a quarter of the CDC has been cut under this administration, and that's going to have ripple effects across all of public health. In this situation, it means fewer people conducting outbreak investigations, which can really hamper the response.\n\nTaken together, it was a week that put technology, politics, policy, and public health head to head all at once. OpenAI is battling challenges in court and inside its own offices, New York has taken a step on data centers whose echoes could reach other states, DOGE's use of AI remains sealed in a black box, and a humble salad on the kitchen table has suddenly become a nationwide worry. These stories may look separate, but at the heart of all of them is the same question, of where the ordinary person fits amid rapidly changing technology and the decisions being made around it.\n\nWhat this means for you\n• Watch what you eat: With cyclosporiasis now in more than 30 states, avoid ordering salads out and scrub your lettuce thoroughly at home, since this parasite survives bleach, chlorine and vinegar.\n• Energy bills and water: People living near large data centers face higher utility bills and heavier water use, so rules like New York's could soon become a live debate in your own state.\n• For tech buyers: The Apple versus OpenAI fight will help decide what new AI devices, beyond the phone, actually reach consumers in the coming years.\n\nQuestions & Answers\n\n1. Why did Apple sue OpenAI?\nApple alleges OpenAI stole sensitive hardware information, including unreleased iPhone parts, prototypes, confidential designs and documents about secret projects. The alleged theft happened mostly through former Apple employees.\n\n2. Why is Tang Tan named in the lawsuit?\nTang Tan is OpenAI's chief hardware officer and spent 24 years at Apple. Apple accuses him of encouraging departing employees to bring proprietary information and unreleased technology with them.\n\n3. What is Guardrails Alliance?\nIt's a super PAC funded by some OpenAI staffers, launched last month with $5 million, that pushes for tighter regulation of frontier AI labs. It aims to raise $15 million during this election cycle.\n\n4. Who does New York's data center moratorium apply to?\nIt applies only to large-scale data centers of 50 megawatts or more and imposes a one-year pause. It does not affect facilities already under construction or holding existing permits.\n\n5. How did DOGE use AI at HUD?\nDOGE members used AI to identify agency rules for potential rescission or contract cancellations. The agency is now denying FOIA requests for information about how those tools were used.\n\n6. Is there an AI exemption under FOIA?\nNo, there is no AI exemption under FOIA, and AI communications are not privileged. There are also no US laws requiring the government to disclose whether AI was used to create rules, policies or regulations.\n\n7. How widespread is the cyclosporiasis outbreak?\nIt has spread to more than 30 states, and the CDC suspects there may be close to 7,000 cases nationwide, with more than 3,700 in Michigan alone.\n\n8. How can you protect yourself from cyclospora?\nScrub all raw vegetables, especially lettuce, even if the bag says pre-washed. Avoid salads when eating out and watch for dehydration, since the parasite survives bleach, chlorine and vinegar.",
  "url": "https://trendkia.com/en/ai/openai-ke-lie-bhari-para-eka-haphta-apple-ka-hardaveyara-chori-ka-mukadama-karmachariyon-ki-bagavata-aura-30-rajyon-men-phailata-k-8245",
  "category": "AI",
  "publishedAt": "2026-07-16",
  "tags": [
    "Apple OpenAI lawsuit",
    "Tang Tan",
    "data center moratorium",
    "Kathy Hochul",
    "cyclosporiasis",
    "DOGE AI",
    "Guardrails Alliance",
    "New York"
  ],
  "language": "en",
  "site": "TrendKia"
}