Who doesn’t enjoy a thrilling case of FOMO? From the dot-com era to Web 2.0, virtual reality to blockchain, the technology sector has seen its fair share of anxiety about missing out on trends.

The AI bubble is the grandest of them all. Its initial offspring — the frenzy to secure power for data centers — is now spawning a frantic race to obtain natural gas resources and machinery. If FOMOs could reproduce, the AI bubble is already in the grandparent phase.

On Tuesday, Microsoft announced that it’s collaborating with Chevron and Engine No. 1 to construct a natural gas power plant in West Texas that could expand to generate 5 gigawatts of electricity. This week, Google verified that it’s partnering with Crusoe to create a 933 MW natural gas power facility in North Texas. Additionally, last week, Meta revealed it would be adding seven more natural gas power plants to its Hyperion data center in Louisiana, escalating the site’s capacity to 7.46 GW — sufficient to power the entire state of South Dakota.

Is there anyone we’ve overlooked?

The recent funding is primarily focused in the southern U.S., which hosts some of the largest natural gas reserves globally. The U.S. Geological Survey recently estimated that in one area alone, there’s enough to provide energy to the entire United States for 10 months. Every data center operator seems eager to stake a claim.

The rush for natural gas has caused a shortage of turbines for power plants, with prices expected to soar by 195% by year’s end compared to 2019 figures, as noted by Wood Mackenzie. This equipment accounts for 20% to 30% of the total cost of a power plant. Firms won’t be able to make new orders until 2028, and turbine delivery is taking an estimated six years, according to the consultancy.

This indicates that tech companies are wagering that the AI craze will persist, that AI will require ever-increasing amounts of energy, and that natural gas power generation will be crucial for triumph in the AI age.

They may eventually regret that third assumption.

While the U.S. boasts ample natural gas supplies, and shipping the fuel isn’t economical, the country is somewhat shielded from the unrest in the Middle East. However, supplies are not infinite, and recently, growth in production from the major three regions — responsible for three-quarters of all U.S. shale gas output — has decelerated significantly.

It’s uncertain how insulated tech firms are from price fluctuations as no specific terms of their agreements have been disclosed. Much will hinge on how firm the pricing is in those contracts.

Even if the contracted prices are as stable as possible, companies could still encounter challenges.

Since natural gas accounts for around 40% of electricity generation in the U.S., according to the Energy Information Administration, electricity costs are closely linked to natural gas prices. Tech firms might briefly mitigate scrutiny by relocating their gas power plants behind the meter — bypassing the grid and connecting them directly to their data centers. Nevertheless, natural gas is a finite resource, and if their aspirations expand too significantly, even behind-the-meter operations could inflate power prices for all. This pattern has been observed before.

It won’t solely be ordinary households that become disgruntled. Other sectors, including those more reliant on natural gas who haven’t yet switched to renewables, might object to data centers monopolizing so much of the resource. Powering a data center with wind, solar, and batteries is straightforward. Operating a petrochemical facility? Not nearly as simple.

Then there’s the factor of climate. A single harsh winter could alter the dynamics by increasing demand among households. Wellheads might freeze, drastically reducing supplies, as experienced in Texas in 2021. When gas becomes scarce, suppliers will be faced with a dilemma: keep the AI data centers operational or allow residents to heat their homes?

By acquiring natural gas resources and operating behind-the-meter, tech companies can assert that they’re “providing their own power” and not straining the electrical grid. However, in reality, they’re merely transferring their usage from one grid to another, the natural gas network. The AI surge has highlighted how physically limited the digital realm remains. Is it wise for them to place big bets on a limited resource? Tech firms may come to regret falling for the FOMO.

As data centers have expanded and multiplied, so has the resistance.

A recent poll from Harvard/MIT indicated that 40% of individuals were in favor of constructing a data center in their vicinity, with 32% against it when queried about the establishment of various industrial sites in their neighborhoods.

One interesting detail from the survey, according to Axios: More individuals would prefer an e-commerce distribution center.

Two-thirds of those surveyed in the 1,000-person poll carried out in November expressed concerns that a new data center in their locality could drive up electricity costs. Enthusiasm for job creation and economic development supported the argument for data centers, as reported by Axios — although that feeling may diminish since most data center operations don’t generate many jobs once they’re established.

Another survey, performed last month and released earlier this week by Quinnipiac University, revealed significantly greater opposition to the construction of data centers. That poll discovered that 65% of Americans were against the establishment of an AI data center in their neighborhood, while only 24% of the 1,397 U.S. adults surveyed endorsed the idea.

The recent polls indicate that the discussion surrounding data centers is far from resolved, and ongoing dissatisfaction from such a significant portion of the electorate is likely to continue influencing politics. Data centers used to operate quietly in the background, but that’s changed.

It’s been on the horizon for weeks, but the conclusion is approaching: Only a few hundred Tesla Model S and Model X cars are left available for sale. Tesla’s CEO Elon Musk confirmed in a post on X this week that custom orders for the Model S sedan and Model X SUV have concluded. “All that remains are a few in inventory,” he stated.

Musk initially revealed Tesla’s intention to halt Model S and Model X production back in January. The data clarifies the reasoning behind this.

Sales of the Tesla Model X and Model S have consistently decreased over the years as the company’s higher volume and more affordable models — the Model 3 and Model Y — have risen in popularity. Tesla aggregates S and X sales under “other models,” which now includes the Cybertruck. The combined data shows S and X sales peaked in 2017 at 101,312 units before dropping to 50,850 units (Cybertruck inclusive) in 2025 — a minute fraction of the 1.63 million vehicles it shipped worldwide last year.

In simple terms, their demise was unavoidable. The subsequent developments are slightly more complex.

Musk is not replacing the void created by the Model X and Model S with a conventional EV; he abandoned plans to manufacture a more affordable EV that was anticipated to cost around $25,000. Instead, Musk is betting on the Optimus robot, which has yet to commence production, and the Cybercab, an entirely electric two-seater autonomous vehicle that was first introduced as a concept in 2024.

Tesla aims to manufacture Optimus robots at its Fremont, California, facility once production of the Model S and Model X concludes, potentially any day now as final orders have been placed. Musk has indicated that Tesla will start producing the Cybercab this month at its factory in Austin, Texas.

A retrospective

The Model S and X EVs have taken a backseat to the more budget-friendly Model 3 and Model Y vehicles. However, their launches and initial sales represented two pivotal moments in Tesla’s vibrant and often tumultuous history. The Model S debuted in 2012 as the company’s first mass-market EV. Its acclaim not only transformed consumer perceptions of EVs but also compelled traditional automakers — long skeptical of the significance of electric vehicles — to pay attention.

The Model X followed in autumn 2015 and was famously dubbed by Musk as the Fabergé egg of EVs.

“I think we got a bit overenthusiastic with the X,” Musk remarked in a September 2015 press interview attended by this reporter just an hour prior to Tesla’s Model X delivery event. “I’m not certain anyone should manufacture this car.”

The Model X was frequently delayed and initially critiqued for its complexity. Nevertheless, it ultimately opened doors for the company to a new demographic: women.

The Model X elevated Tesla’s visibility, positioning the company for its next significant venture: an affordable mass-market EV. The Model 3 experienced a rocky beginning, but ultimately propelled Tesla into the spotlight. The Model Y solidified its status, helping Tesla to extend its lead as the top-selling EV manufacturer worldwide until China’s BYD claimed that leading global EV sales position in 2025, delivering 2.26 million EVs.

Tesla continues to sell thousands of Model 3 and Model Y units, but its growth has stagnated and even reversed. The company announced in January that it sold 1.69 million vehicles in 2025, a decline for the second consecutive year. Its attempts to boost sales with more economical, stripped-down versions of the Model 3 and Model Y launched in October have shown some degree of success, based on first-quarter 2026 numbers reported on April 2.

In the first three months of the year, Tesla delivered 358,023 EVs globally, approximately 6% more than during the same timeframe in 2025, which was the company’s worst quarter in years. This figure fell short of analysts’ projections of around 368,000.

But that is of little concern. In Musk’s perspective — one for which he is well rewarded — Tesla is not merely an automaker or a sustainable energy entity, as he has previously stated. Tesla is fundamentally an AI company, and his new strategy fully embraces that vision.

Cybercab challenges

The Optimus robot is one component of Tesla’s AI initiative. However, it is the Cybercab that most clearly captures and unveils the risks associated with the company’s AI-centric approach.

The Cybercab is engineered to operate as an autonomous vehicle without conventional controls such as a steering wheel or pedals, meaning it will launch without the support of a human safety operator.

The inaugural Cybercab exited the Tesla factory assembly line in February and is scheduled to begin mass production this month. However, that timeline may be pushed back, as many have in Tesla’s past.

Unlike Tesla’s earlier vehicles, the challenges here do not lie in the production (who can forget the production struggles of the Model 3?). Instead, it faces significant regulatory impediments before it can hit the streets. Federal motor vehicle safety standards require vehicles to include a steering wheel and pedals. There is no indication that Tesla has sought an exemption, based on publicly available records from the Federal Register and the National Highway Traffic Safety Administration.

The vehicles will also depend on Tesla’s Full Self-Driving software to navigate public thoroughfares and securely transport passengers to their destinations. Despite advancements in FSD and limited trials with driverless robotaxis in Austin, Tesla has yet to prove that its software can function reliably at scale.

And that aspect demands more than just technical expertise. Robotaxi operations are also complex. In places like California, they require permits to deploy and charge for rides in driverless vehicles.

Zoox, the autonomous vehicle firm owned by Jeff Bezos’ Amazon, could pave the way for Tesla and its Cybercab. Zoox received an exemption from the National Highway Traffic Safety Administration permitting it to showcase its custom-designed robotaxis, which do not feature pedals or a steering wheel, on public streets. Zoox is currently undergoing a public process to have that exemption expanded to commercial activities.

Musk attempted to convince shareholders about the merits of the risk during the company’s earnings call in January.

“The vast majority of miles traveled will be autonomous in the future,” Musk stated at the time, adding that the Cybercab is highly optimized for the lowest cost per mile and also for a significantly higher usage rate. “I would estimate probably less than, I’m just hypothesizing, but probably less than 5% of miles driven will see someone actually driving the car themselves in the future, perhaps as low as 1%.”

On Thursday, the EU’s cybersecurity agency announced that a recent hacking incident and data breach affecting the EU’s executive arm was carried out by a cybercriminal organization identified as TeamPCP. 

In a fresh report, CERT-EU disclosed that the attackers managed to steal approximately 92 gigabytes of compressed data from an Amazon Web Services (AWS) account utilized by the European Commission, which included sensitive information such as names, email addresses, and email content. 

The breach impacted the cloud infrastructure of the Commission’s Europa.eu platform, a site leveraged by member states for hosting the websites and publications of various EU institutions and agencies.

CERT-EU indicated that the data of at least 29 other EU entities could be compromised, and numerous internal clients of the European Commission might also have had data extracted. 

Subsequently, the stolen information was published online by another hacking entity, the infamous ShinyHunters. 

While the scale of the data breach is indeed significant, it is uncommon for a cybersecurity agency to link two distinct hacking groups to the same event. A representative from ShinyHunters informed TechCrunch in a chat that they had secured some of the data initially taken by TeamPCP in preceding attacks, which they then leaked.

Attempts to contact TeamPCP for a statement were unsuccessful.

CERT-EU revealed that the breach began on March 19 when hackers obtained a confidential API key related to the European Commission’s AWS account, stemming from a prior hack on the open source security tool Trivy. The Commission unintentionally downloaded a version of the compromised Trivy tool after the project’s recent breach, allowing the hackers to access its secret API key, which enabled them to retrieve data stored in the Commission’s AWS account.

While the agency continues to analyze the data that was released online, nearly 52,000 files contain sent email messages. CERT-EU indicated that most of these emails are automated with minimal content, but emails that returned with an error “may contain the original user-submitted content, creating a risk of personal data exposure.”

CERT-EU stated it is already reaching out to the organizations impacted. 

A representative for the European Commission informed TechCrunch that the agency is closed until next week and will provide a response to requests for comment then. 

In addition to the Trivy breach, TeamPCP has been associated with ransomware assaults and cryptocurrency mining operations, according to Aqua Security, the developer of Trivy. Recently, the hackers have been linked to a systematic series of supply chain attacks compromising additional open source security initiatives, as reported by Palo Alto Networks Unit 42.

By targeting developers with keys that grant access to sensitive systems, the hackers “then have the capacity to extort compromised organizations for ransom, demanding payments,” wrote Unit 42.

This article was revised to encompass comments from a member of ShinyHunters.

When Brett Levenson departed Apple in 2019 to oversee business integrity at Facebook, the social media platform was amid the aftermath of the Cambridge Analytica scandal. Initially, he believed he could rectify Facebook’s content moderation issues through enhanced technology. 

However, he soon discovered that the issue was more profound than just technology. Human reviewers were tasked with memorizing a 40-page policy document that had been translated into their native language by a machine, he explained. They then had approximately 30 seconds for each piece of flagged content to determine not only if it breached the rules but also what action to take: block it, ban the user, or limit its distribution. According to Levenson, those rapid decisions were only “slightly better than 50% accurate.”

“It was almost like flipping a coin to see if the human reviewers could correctly interpret the policies, and this was after several days of the harm already occurring,” Levenson stated to TechCrunch.

Such a delayed, reactive method is not viable in a landscape filled with agile and well-funded adversaries. The emergence of AI chatbots has only amplified the situation, as failures in content moderation have resulted in a series of notable incidents, including chatbots directing teens toward self-harm or AI-generated images bypassing safety filters.

Levenson’s dissatisfaction prompted the concept of “policy as code” — a method to convert static policy documents into actionable, updatable logic closely tied to enforcement. This insight led to the establishment of Moonbounce, which has disclosed a $12 million funding raise on Friday, as learned exclusively by TechCrunch. The funding round was co-led by Amplify Partners and StepStone Group.

Moonbounce collaborates with businesses to add an extra layer of safety wherever content is produced, whether by a human or by AI. The company has developed its own large language model to examine a client’s policy documents, assess content in real-time, provide a response in 300 milliseconds or less, and take appropriate action. Depending on the client’s preferences, this action might involve Moonbounce’s system delaying distribution while the content awaits a human review or could result in blocking high-risk content immediately. 

Currently, Moonbounce caters to three primary sectors: platforms handling user-generated content such as dating applications; AI companies creating characters or companions; and AI image generation services. 

Moonbounce is facilitating over 40 million daily reviews and caters to more than 100 million daily active users on the platform, stated Levenson. Clients include AI companion startup Channel AI, image and video generation firm Civitai, and character roleplay platforms Dippy AI and Moescape. 

“Safety can actually serve as a product advantage,” Levenson remarked to TechCrunch. “It has never been perceived this way before because it has always been an afterthought, not something that can be integrated into the product. We observe that our clients are discovering fascinating and innovative methods to leverage our technology to make safety a competitive edge and part of their product narrative.”

Tinder’s trust and safety lead recently discussed how the dating service utilizes these types of LLM-powered solutions to achieve a tenfold improvement in detection accuracy.

“Content moderation has always been a challenge for major online platforms, but with LLMs now central to every application, this challenge has become even more formidable,” commented Lenny Pruss, general partner at Amplify Partners, in a statement. “We chose to invest in Moonbounce because we envision a future where objective, real-time guardrails become the essential infrastructure of every AI-mediated application.”

AI enterprises are under increasing legal and reputational pressure after chatbots have been criticized for guiding teenagers and vulnerable users towards suicidal thoughts, with image generators like xAI’s Grok being used to create non-consensual nude images. Clearly, internal safety measures are failing, and it’s turning into a liability issue. Levenson mentioned that AI companies are progressively seeking assistance from external sources to enhance their safety frameworks. 

“We act as a third party positioned between the user and the chatbot, which means our system isn’t overwhelmed with context in the same way the chat is,” Levenson explained. “The chatbot must remember, possibly, tens of thousands of tokens that have been exchanged previously…Our sole concern is enforcing the rules in real-time.”

Levenson co-manages the 12-person company with his former Apple associate Ash Bhardwaj, who previously developed large-scale cloud and AI infrastructure across Apple’s core products. Their upcoming focus is a feature known as “iterative steering,” created in response to incidents like the 2024 suicide of a 14-year-old boy from Florida who became fixated on a Character AI chatbot. Instead of an outright refusal when harmful subjects come up, the system would intervene in the conversation and reroute it, adjusting prompts in real time to steer the chatbot toward a more actively supportive response.

“We hope to incorporate into our action toolkit the capability to guide the chatbot in a more positive direction, effectively modifying the user’s prompt to compel the chatbot to be not only an empathetic listener but also a helpful listener in such situations,” Levenson stated. 

When queried about whether his exit strategy involved an acquisition by a company such as Meta, completing his journey with content moderation, Levenson acknowledged how well Moonbounce would integrate into his former employer’s offerings, as well as his responsibilities to his investors as a CEO. 

“My investors would be furious if they heard me say this, but I would despise seeing someone purchase us and then limit the technology,” he expressed. “Like, ‘Alright, this is ours now, and no one else can benefit from it.’”

The conflict in Iran has severely impacted global oil markets, leading to a sharp increase in gas prices in the U.S. In light of the escalating transportation expenses, Amazon has implemented a new 3.5% fuel surcharge for sellers utilizing its distribution network. This policy could result in substantial additional costs for the countless merchants relying on the e-commerce giant to market their products.

Amazon informed TechCrunch that the surcharge would remain in effect for the foreseeable future, though the company stated it would keep assessing a potential adjustment to the policy as market circumstances continue to change. The announcement was initially reported by Bloomberg.

“Rising costs in fuel and logistics have heightened operational expenses throughout the industry,” a representative mentioned. “We have managed to absorb these increases thus far, but akin to other leading carriers, when costs stay high, we impose temporary surcharges to partially recoup these expenses.” The representative added that the surcharge is “notably lower than those charged by other major carriers.”

The updated policy will go into effect on April 17 and will affect sellers using the company’s Fulfillment by Amazon service, as reported by Bloomberg. Fulfillment by Amazon, or FBA, enables businesses to send their goods to Amazon’s warehouses, where they are packaged and dispatched to customers. Amazon does not reveal how many merchants utilize FBA, but the program supports the vast majority of third-party transactions on its platform.

Amazon first established this type of surcharge in 2022—which, not entirely coincidentally, was the last occasion crude oil prices exceeded $100 a barrel. What was unfolding in 2022? Russia had just invaded Ukraine, causing chaos in energy markets. Currently, the conflict in Iran—triggered by actions from the Trump administration and the assassination of the nation’s Supreme Leader by the Israeli government—has similarly disrupted markets.

Iran is strategically positioned along the northern edge of the Strait of Hormuz—a narrow yet vital shipping route for global oil supplies through which approximately 20% of the world’s oil supply flows—and the nation has attempted to obstruct shipping lanes in that region, causing a significant impact on energy prices globally.

Hims & Hers, the telehealth firm offering weight-loss medications and sexual health prescriptions, has acknowledged a data breach impacting its external customer service system.

The healthcare organization stated in a breach notification submitted to the California attorney general on Thursday that cybercriminals accessed data regarding user inquiries directed to the company’s support team. The firm reported that hackers infiltrated its third-party ticketing platform between February 4 and February 7, resulting in the theft of numerous support tickets containing personal information provided by customers.

According to the data breach notice, the hackers acquired customer names and contact details, along with other unspecified personal information that Hims & Hers opted to redact in the correspondence.

While the company claims that customer medical records were unaffected, the nature of support systems implies that the data may include sensitive details regarding an individual’s account, personal information, and healthcare.

It remains unclear how many individuals had their personal information compromised in the incident. California law mandates that organizations disclose data breaches affecting 500 or more residents of the state.

Jake Martin, a representative for Hims & Hers, informed TechCrunch in a statement that the company experienced a social engineering attack, where hackers deceive employees into providing access to their systems. The spokesperson mentioned that the stolen data “mainly consisted of customer names and email addresses.” The company did not specify the exact types of data that were taken when asked by TechCrunch.

The company declined to comment on whether it has received any communications from the hackers, including ransom demands.

Recently, customer support and ticketing systems have become lucrative targets for financially motivated cybercriminals, who have compromised databases containing customer information and coerced companies into paying ransoms.

Last year, Discord experienced a data breach that impacted its customer support ticketing platform and revealed government-issued IDs of approximately 70,000 individuals who had submitted their driver’s licenses and passports to verify their age.