The big picture: The fight over data centers is increasingly centered on what these facilities could mean for nearby residents’ power bills, water supplies, and quality of life. New research suggests that opposition is driven less by fears about AI itself than by concerns that communities will be left to absorb the infrastructure costs required to support it.

Researcher Andy Masley reviewed polling data on the backlash and concluded that its main drivers are not simply distrust of Big Tech, skepticism about AI, or concerns about China. Instead, he found that people are responding to “object-level local environmental and economic harms” associated with proposed projects.

Power is often the first concern. Large data centers can require significant grid upgrades, including new substations, transmission lines, and other equipment. In the PJM Interconnection region, those costs have often been spread across all customers, even when the upgrades are driven by data center demand. Some customers have seen their utility bills rise by as much as 76%.

The issue has become significant enough to draw attention from the Trump administration. President Trump brought together technology companies, AI hyperscalers, utilities, and governors to sign a “ratepayer protection pledge” intended to ensure that AI companies “pay their own way” for the energy infrastructure they require.

Some states are beginning to address the issue directly. Oregon has adopted a law requiring facilities that use at least 20 megawatts of electricity to cover their share of grid-related costs. Portland General Electric, the state’s largest utility, raised rates for data centers by 30% while cutting residential rates by 1.3%.

Water use has also become a flashpoint. In Georgia, reports linked data center operations to low water pressure and, in one case, muddy water coming from household taps. A separate dispute in Wyoming involved a Meta contractor that allegedly contaminated Cheyenne’s reclaimed-water system. City officials later suspended certain fill-and-flush and closed-loop discharges associated with the project.

Other communities have raised concerns about noise and emissions. Residents near a Michigan data center said the facility produced a constant sound, comparing it to having “someone set up a vacuum, like in your living room.” The company was fined for violating a local industrial noise ordinance and offered to buy homes from nearby residents.

Some projects have also faced criticism over the use of mobile turbine generators or plans to build dedicated gas-fired power plants. A proposed Amazon-related facility in Texas includes a 7.65-gigawatt natural gas plant authorized to emit up to 33 million tons of greenhouse gases annually.

Masley argues that these are concrete concerns, not vague fears. They stem from issues that residents can see, hear, or pay for. However, he also noted that coverage of individual disputes can sometimes leave out important details about proposed projects or fail to compare them with other data center sites. That can lead people to assume that problems reported elsewhere will automatically occur in their own communities.

For developers, this makes transparency more important. Residents want to know how much power a site will use, who will pay for grid upgrades, where the water will come from, and how it will be handled after use. They also want clear information about backup generators, emissions, and noise controls.

Instead, some developers have taken a more confrontational approach. Several have sued local governments that imposed data center bans or moratoriums, arguing that officials exceeded their authority or violated due process and equal protection rules.

In one Pennsylvania town, officials presented a developer with 43 requirements before approving a project. The developer considered the demands too onerous, withdrew from the talks, and later reapplied. The council described the move as an attempt at “approval by tantrum.”

The dispute over data centers is unlikely to disappear as demand for AI computing capacity grows. However, the resistance is not only about the technology itself. It is also about whether the costs of building that infrastructure will be borne by the companies using it or by the communities living next to it.

Image credit: Andy Masley

Presented by Nutanix

Autonomous systems that can reason, make their own decisions, and execute actions across an environment introduce a category of risk that application-level controls were never built to contain. Treating that risk as a single problem produces incomplete architectures, says Oscar Wahlberg, senior director of product management at Nutanix.

“The guardrails to catch a malicious prompt won’t stop an agent from hallucinating and doing something it never should have done, like accidentally deleting databases or leaking sensitive data with a credential it was granted but then uses for something entirely different,” Wahlberg says. “That’s the central problem as enterprises move autonomous agents out of experimentation and into production.”

Once an agentic system is granted execution privileges across the data center, the security posture has to scale into a defense-in-depth architecture spanning infrastructure, storage, compute, networking, and a governing control plane. Each layer addresses a distinct category of risk, rather than duplicating the same controls across the stack. No single security control or vendor can provide that protection on its own. Defense-in-depth depends on those layers working together.

By dividing the responsibilities across the stack and adhering to zero trust segmentation, organizations can create a secure framework that improves their overall posture. Understanding which risks belong in each layer is what turns the principle of defense-in-depth into a practical security framework, with three layers that each have a distinct responsibility.

Infrastructure layer: Establishing trust where AI agents run

The infrastructure layer’s foundational responsibility is establishing a root of trust that answers a simple question: who is operating in the environment? That trusted identity becomes the prerequisite for every security control above it. Before an organization can trust what an agent does, it first has to trust the integrity of the environment where the agent runs. When an agent requests permission to execute an operation, the system must be able to verify that the request came from the legitimate agent — not something impersonating it.

Delivering that kind of assurance depends on technologies that root trust in the hardware itself, including platform attestation, confidential computing, and secure boot, alongside controls that prevent unauthorized access both within a server and beyond it. For regulated industries such as financial services, this layer provides the ability to isolate AI production workloads so that neither the agent nor the environment can operate outside its assigned scope. That mitigates risks including model and runtime tampering, supply chain compromise, and unauthorized access to sensitive AI workloads.

Network layer: Governing how AI agents communicate

Once agents begin communicating with other agents, APIs, applications, and enterprise systems, they generate a level of concurrency and dynamic communication that traditional static network configurations were never designed to handle. An agent configured to call APIs, query data sources, and spin up additional agents without constraint creates a sprawling web of east-west traffic that becomes very difficult to reason about, and that complexity can easily mask lateral movement or data exfiltration when the right network security layers are not in place.

“We should treat AI agents as a new class of network identity, and make sure that an agent can only talk to other agents or data sources where it’s explicitly allowed to do so,” Wahlberg says. “That means moving away from rigid static rules toward dynamic policy enforcement.”

Nutanix’s solution is Agent Gateway, part of the Nutanix Agentic AI solution. It’s a unified, governed layer that is designed to provide cost control and governance capabilities to help manage autonomous agent users. Coupled with agents grounded in zero trust segmentation and using capabilities like Nutanix Flow for micro segmentation and integrating with networking vendors, including its integration into the Cisco Secure AI Factory, Agent Gateway helps enterprises govern interactions across agents, models, data sources, and enterprise applications.

The network layer governs lateral movement, data exfiltration, and gates the agent’s network interactions. A zero trust framework with access blocked by default and scalable interaction monitoring is important for agents since they can exhibit unreliable behavior. The Nutanix software integration with Cisco UCS servers and Cisco AI PODs delivers the turnkey physical infrastructure (compute, storage, and networking) that the AI factory runs on.

Control plane layer: Governing what AI agents are permitted to do

The control plane is the brains of the operation, providing a central point for managing agent permissions, tool access, resource consumption, and runtime visibility. What matters most is having a single place where policies can be enforced consistently rather than reinvented for every agent, Wahlberg says.

“Agent Gateway acts as a universal endpoint for different models and tools, so an IT team can configure their agents to talk to this single control point,” he explains.

The centralized AI gateway enables the admin to observe, audit, and control access to models as well as MCP tools protecting data and gating privileged access. This layer is designed to help mitigate risks such as privilege misuse, runaway agents, unauthorized tool usage, data leakage, and the excessive model consumption that can lead to increased token consumption when agents get stuck in runtime loops. And it depends on treating governance as a runtime control system rather than a compliance afterthought.

Why one-size-fits-all security fails agentic AI environments

The biggest architectural mistake enterprises make is assuming a single security model can be stretched across every layer of an AI stack. When an organization tries to solve for hardware-level trust with application-level software, or leans on static legacy network rules to manage dynamic agents, it builds an architecture that either blocks the agentic system from doing its job or leaves critical doors wide open. One-size-fits-all thinking tends to produce significant performance penalties and operational friction.

“By failing to assign specific responsibilities to the appropriate layers, enterprises end up with blind spots in governance,” Wahlberg says. “They might secure the model output but miss that there’s data leakage between agents, or they might secure the network but lack the control plane visibility to understand that they’re wildly burning tokens because the agents are stuck in some kind of runtime loop.”

Focusing exclusively on the model leaves the largest gaps of all, because a guardrail that catches a malicious prompt does nothing to stop a hallucinating agent from misusing a legitimate credential. Embedding security across the full stack helps ensure that even when a model level threat slips past the initial filters, the agent remains constrained by hardware rooted trust, network isolation, and access controls at the agent layer.

How Intel, Cisco, and Nutanix build defense-in-depth together

The three-way partnership from the three companies demonstrates how the layered architecture comes together in practice as a well-governed, enterprise-grade AI Cloud. Intel supplies the computer to run agentic workloads and secures the execution environment through hardware-rooted trust and confidential computing, while also driving costs down through their accelerators. Intel Xeon 6 processors with built-in AMX accelerate AI inference efficiently without relying exclusively on expensive GPUs.

Cisco wraps the environment in a secure fabric that governs communication between agents and enterprise tools, while Nutanix provides the software platform, minimizing architectural silos, and the central control plane that enforces permissions, delivers visibility and cost governance, and ties the architecture together into a defense-in-depth solution that lets enterprises scale agentic AI.

Of the three layers, enterprises currently underestimate the control plane the most, Wahlberg says. A true control plane extends far beyond initial deployment to simplify Day 2 operations, he explains, giving IT teams the continuous observability, and strict token governance required to keep autonomous agents secure and cost-effective in production.

“Apart from model and tool selection, governing the agent deployments and their access to models and business tools in a tightly integrated full stack platform will be important for the success of AI projects,” he says, pointing to a near future in which organizations move from a handful of AI use cases to thousands of agents working autonomously to drive the business.

Technology leaders should prioritize building a centralized governance layer today that can manage agent identities, tool permissions, and token budgets in real time, because that control point is what builds the operational muscle to scale safely.

“You can’t build an AI system without getting into a lot of complex decisions,” he explains. “And you need a control plane that talks across multiple vendors and infrastructures to help you solve for those defense-in-depth strategies.”

Learn more about the Nutanix Agentic AI solution here.

Sponsored articles are content produced by a company that is either paying for the post or has a business relationship with VentureBeat, and they’re always clearly marked. For more information, contact sales@venturebeat.com.

A Kentucky-based glass plant is poised to become a much bigger piece of the iPhone supply chain, with hundreds of both long-term and temporary jobs expected as production ramps up.

Apple has been getting glass for iPhone for 20 years from Kentucky. Specifically, in Harrodsburg, Kentucky, where the tech giant plans to expand manufacture the glass for every iPhone and Apple Watch.

Apple COO Sabih Khan and U.S. Secretary of Commerce Howard Lutnick toured Corning Inc. on Friday. The tour comes one year after Apple had invested $2.5 billion more in Kentucky, hoping to double plant jobs and triple its current output.

“Made in Kentucky. Made in America. Sold around the world,” U.S. Rep Andy Barr said during the tour, according to The Lexington Herald-Leader. “We invent it here. We have the talent, and we have the manufacturing heritage. We will build it here, and we will sell it to the world.”

“I want Kentucky to be the first place companies think about when deciding to make their next major American investment.”

The plant currently has 350 employees, but hopes to employ 200 new hourly production jobs, alongside a handful of salaried employees in glass development. That’s not quite double the existing 350, but that’s because Corning was including construction jobs in that total.

However, Apple and Corning are planning to build a new Apple-Corning Innovation Center at the site. It is believed that construction will create 100 jobs, but those won’t last.

“At Apple, we share the administration’s commitment to producing more of the most innovative technology right here in the United States,” said Khan.

“We’re expanding our U.S. supply chain, and we will continue to build on these efforts because at Apple, we believe in the power of American innovation. We believe in the ingenuity of American workers, and we believe deeply in the promise of America.”

The Apple-Corning Innovation Center may be similar to Apple’s Advanced Manufacturing Center in Houston, though likely at a smaller scale.

If you are stuck with the same five prestige dramas everyone’s already talking about, here are three that got criminally overlooked. For this weekend, we have a yakuza thriller, a finance drama, and a televangelist saga that still delivers after four seasons. All three underrated shows are streaming right now on HBO Max, and deserve a spot on your watchlist. If you don’t have the time or the appetite to commit to multiple seasons, I have created a list of the best limited series on Netflix that’s worth checking out.

We also have guides to the best new movies to stream, the best movies on Netflix, the best movies on Hulu, the best free movies, and the best movies on Amazon Prime Video.

Tokyo Vice (2022-2024)

Genre: Crime, Drama, Thriller
IMDb: 8.0/10
Rotten Tomatoes: 89%

Jake Adelstein (Ansel Elgort) talks his way onto a Japanese newspaper’s crime desk in the late ’90s, then keeps pulling threads until he’s tangled up with actual yakuza. Detective Hiroto Katagiri (Ken Watanabe) half mentors him, half tolerates him, and their partnership carries the whole show. Underneath the neon and the subtitles, this is really a story about how far someone can go for a byline.

Michael Mann directed the pilot, setting a gorgeous, moody, authentic tone that you can feel in every frame. The secret weapon here is Show Kasamatsu, who plays a rising mobster with incredible charisma. This show on HBO Max has two tight seasons, zero filler, and an ending that actually reaches a resolution.

You can stream Tokyo Vice on HBO Max

Industry (2020- present)

Genre: Drama
IMDb: 7.5/10
Rotten Tomatoes: 91%

Harper Stern (Myha’la Herrold) and Yasmin Kara-Hanani (Marisa Abela) fight for a handful of permanent jobs at a London investment bank, and every day on the trading floor is an absolute knife fight for survival. They drown their grueling late-night hours in substance abuse, complex office politics, and deeply terrible relationship choices. The show is also about what ambition does to people who never learn to stop.

What got me hooked wasn’t the finance jargon; it was watching characters make self-destructive choices and finding myself still rooting for them. I’ve recommended this to friends who work in soul crushing jobs, and they texted me mid-season one saying, “this is too real.” Four seasons in and a fifth already confirmed, this thing keeps getting sharper instead of running out of ideas.

You can stream Industry on HBO Max

The Righteous Gemstones (2019-2025)

Genre: Comedy, Drama
IMDb: 8.0/10
Rotten Tomatoes: 91%

Eli Gemstone (John Goodman) tries to keep his three deeply flawed adult children from tearing apart the megachurch empire he built. Jesse (Danny McBride), Judy (Edi Patterson), and Kelvin (Adam DeVine) spend four seasons scheming against each other and occasionally the whole family at once. It’s a dark comedy about how faith and greed end up looking identical once enough money gets involved.

I put off watching this for years, assuming it was just Danny McBride yelling for four seasons, but the show proved me wrong. He channels pure satirical genius here, delivering funny moments without dropping the narrative. There is also a real tenderness buried under all the chaos, which I liked. And this hidden gem on HBO Max wrapped up with a perfect critic score on the finale.

You can stream The Righteous Gemstones on HBO Max

Your email address is one of the easiest ways for websites and advertising companies to identify you online. If you provide the same address to multiple websites, it can potentially be used to connect your activity across different services.

Brave is now trying to limit this type of tracking. Starting with desktop version 1.94, the browser has added Email Aliases, which let users provide websites with a different email address instead of revealing their primary one.

How do Brave Email Aliases work?

Once users create a Brave Account, they can generate an alias directly from an email field on a website. Any messages sent to this address are then forwarded to the primary email address connected to the Brave Account.

The website never receives the primary address. Companies can upload customer email addresses to advertising platforms such as Google, Meta, and LinkedIn to match them against existing user profiles. Since this matching can happen on company servers, browser-based tracker blocking cannot always prevent it. Aliases can also limit how widely a primary email address spreads if a website suffers a data breach.

However, Brave Email Aliases does not automatically block newsletters or promotional messages. Anything sent to the alias can still be forwarded to the primary inbox. If an alias starts receiving unwanted emails, users can deactivate it, preventing future messages sent to that address from reaching them. Brave is initially offering five aliases for free, while a Premium version and mobile support are planned for the future.

Brave is not the first browser to do this

Email aliases themselves are nothing new. DuckDuckGo Email Protection already provides private @duck.com addresses that forward messages to an existing inbox. It can also remove supported email trackers before forwarding them. Firefox Relay provides a similar service and can generate email masks through Firefox. Its free tier currently supports up to 50 masks.

Brave’s advantage here is having another privacy feature built directly into its browser. The company recently added Containers to separate accounts, cookies, and browsing sessions. Email Aliases now give users another way to keep information shared with different websites separated.

Tim Draper is selling his island in Lake Tanganyika for $7.9M or best offer, announcing it himself on X and inviting offers directly by email. The ECB reported this month that EU venture capital funds hold around 150B euros against roughly 930B euros in the United States.

Tim Draper is selling his island in Lake Tanganyika, Tanzania, for $7.9M or best offer. “Gorgeous place, but we don’t use it enough,” the venture capitalist wrote on X on Friday.

There is no broker and no listing. He is taking offers at his own work address.

The location is not the usual trophy. Lake Tanganyika is the longest freshwater lake in the world and the second deepest after Baikal, and it is shared by Tanzania, the Democratic Republic of Congo, Burundi and Zambia.

Draper is commonly described as a SpaceX and Tesla investor. Both were positions at Draper Fisher Jurvetson, which he co-founded and where his partner Steve Jurvetson led the deals.

The cheques most often called his are elsewhere. Skype, Hotmail, Baidu and Coinbase all sit on his own list.

His best known purchase was not a company at all. He won nearly 30,000 bitcoin at a US Marshals auction in 2014, worth roughly $19M at the market price that week, and has said he did not sell.

Which makes the asking price modest by his standards. It is a reminder of what one early cheque is worth twelve years later, and of who gets to write them.

Europe published its own number this month. EU venture capital funds hold about 150B euros against roughly 930B in the United States, the ECB found, around six times as much.

The shortfall is not even. It widens in later rounds, where companies need the largest cheques.

European scale-ups lean on money from outside the bloc. The ECB flags the risk that successful startups move to where the capital is.

The most uncomfortable finding is about companies nobody funds. Among firms without venture backing, the median EU firm is nearly twice the size of its American equivalent by employment and grows about 15 percentage points faster. Brussels published a Startup and Scaleup Strategy last year promising to widen the pipe.

The composition has not shifted either. The share of European venture going into software and IT services has stalled since 2020.

The people inside those companies own less of them. Employees at late-stage European startups hold roughly half the equity their American counterparts do.

None of which is Tim Draper’s problem. He is selling an island for less than half what one lot of bitcoin cost him in 2014, and Europe’s founders are short 780B euros.