A South Korean semiconductor engineer has leveled a pair of pointed accusations at Tesla, claiming the company pressured Samsung Foundry over pricing for its upcoming AI5 and AI6 chips while a Tesla-linked fabrication project allegedly recruited away Samsung’s own staff. The claims, posted on X under the handle @TTHUEONE and reported by TweakTown, landed on October 18, 2026, and quickly spread across tech and financial outlets tracking the increasingly tense relationship between Tesla and the foundries building its custom silicon.
Neither Tesla nor Samsung has publicly confirmed or denied the allegations, and the specific contract terms, prices, and figures behind them remain unverified. But the episode has reignited scrutiny of how Tesla negotiates with the foundries it depends on for AI5 and AI6, the chips expected to power its next generation of Full Self-Driving computers, Optimus robots, and data center inference workloads.
What the Korean Engineer Actually Alleged
The allegations center on two separate but related claims. First, the engineer said Tesla sought unusually aggressive terms from Samsung Foundry for manufacturing the AI5 and AI6 chips, including pushing for steep price cuts while also demanding access to Samsung’s internal fabrication information and the right to audit Samsung’s operations. Second, the engineer claimed that Terafab, a Tesla-linked fabrication project, had recruited a significant number of employees away from Samsung’s facility in Taylor, Texas.
According to the reporting picked up by Wccftech, the engineer described Tesla’s requested contract terms on AI5 and AI6 as “shitty contract terms [on AI5 and AI6 chips], especially requiring internal fab info sharing and even audits, while beating down the prices like crazy.” On the staffing question, the same engineer said Terafab had poached “a ton of local talent” from Samsung’s Taylor operation.
Both claims are, at this stage, allegations from a single source rather than confirmed facts. No pricing figures, contract values, or headcount numbers tied to the alleged recruitment have been independently verified. Tesla and Samsung have not issued statements addressing the specifics, and the reports circulating since the claims surfaced treat them as unconfirmed pending any official response.
The Companies Caught in the Middle
Five names recur throughout the reporting: Tesla, Samsung, Samsung Foundry, TSMC, and Terafab. Each occupies a distinct position in the dispute, and understanding their roles helps explain why the allegations resonate beyond a single social media post.
| Entity | Role in the Dispute | Alleged Connection |
|---|---|---|
| Tesla | Chip designer and foundry customer | Accused of demanding steep AI5/AI6 price cuts plus internal fab data and audit rights |
| Samsung | Parent company of the foundry unit named in the claims | Alleged target of the pricing and information-access pressure |
| Samsung Foundry | Chipmaking division operating the Taylor, Texas facility | Said to face the contract demands; employees allegedly recruited away |
| Terafab | Tesla-linked fabrication project | Accused of poaching staff from Samsung’s Taylor facility |
| TSMC | Rival contract chipmaker | Named as an alternative manufacturing partner for Tesla’s AI chips |
The reports note an unconfirmed claim that Tesla turned to Samsung Foundry for AI5 and AI6 production partly because TSMC would not meet Tesla’s demands on terms. That detail has not been independently verified by either company and should be read as part of the broader allegation rather than an established fact.
AI5 and AI6: The Chips Behind the Dispute
AI5 and AI6 are the next generations of Tesla’s custom silicon, designed to succeed the AI4 chip currently used in Tesla’s Full Self-Driving computer. Tesla has previously signaled that these chips are meant to handle increasingly demanding workloads, from in-vehicle autonomous driving computation to the processing requirements of its Optimus humanoid robot program and broader data center inference tasks.
Tesla CEO Elon Musk has previously framed the AI5 chip in ambitious terms, describing a performance target of “around 3x the performance of a comparable chip from NVIDIA but at just 10 percent of the cost,” a claim cited in coverage of the Samsung foundry dispute. That kind of target, if accurate, would explain why Tesla might push hard on manufacturing costs with any foundry partner: a chip pitched as dramatically cheaper than Nvidia’s offerings only works financially if Tesla’s own production costs stay low.
Beyond that performance claim, the reports reviewed for this story do not include verified specifications, benchmark results, or production timelines for AI5 or AI6. Any numbers circulating beyond what Tesla or Samsung have officially confirmed should be treated with caution.
The Pricing Squeeze Allegation, Explained
At the heart of the first allegation is a familiar tension in the chip industry: large customers pushing foundries for better pricing, while foundries push back to protect margins on capital-intensive manufacturing. What makes this allegation notable is the claim that Tesla went further than price negotiation alone, reportedly asking for internal fabrication information and audit rights as part of the AI5 and AI6 arrangement.
Audit rights and information-sharing requirements are not unheard of in large semiconductor supply agreements, particularly when a customer wants assurance over yield, quality control, or capacity allocation. But the engineer’s characterization suggests the scope requested by Tesla went beyond what Samsung considered standard, prompting the public pushback. Without access to the actual contract language, outside observers cannot confirm how far the requested terms extended or whether they were ultimately accepted, modified, or rejected.
This dispute surfaces against a backdrop of broader friction between Tesla and its memory and chip suppliers. Elon Musk’s own public remarks about memory manufacturers, covered in Tesla-adjacent reporting on Musk’s memory pricing comments, have already drawn attention to how Tesla talks about supplier pricing. The Korean engineer’s allegation effectively turns that scrutiny back onto Tesla, suggesting the company applies similar pressure to its own manufacturing partners.
The Poaching Allegation and Samsung’s Taylor, Texas Fab
The second allegation concerns Terafab, a project linked to Tesla, which the engineer said recruited a substantial number of employees from Samsung’s semiconductor facility in Taylor, Texas. Samsung’s Taylor site has been one of the most closely watched new fabrication facilities in the United States, and any significant departure of trained staff would be a sensitive issue for a foundry ramping up advanced manufacturing operations.
Talent recruitment between semiconductor companies is common and, in most cases, unremarkable. Engineers move between foundries, fabless designers, and equipment makers constantly, and that mobility is part of how the industry’s institutional knowledge spreads. What elevates this particular claim is the timing and proximity: if Terafab is positioned near Samsung’s existing workforce pool while Tesla is simultaneously negotiating foundry terms with Samsung, the appearance of double-dealing becomes harder for Samsung to ignore, regardless of whether the recruitment was improper.
No confirmed number of recruited employees, job titles, or departure timeline has been published. The claim remains, as of this writing, an assertion from a single social media source rather than a documented personnel count.
Tesla’s Long History of Building Its Own Silicon
This dispute did not emerge in a vacuum. Tesla has spent close to a decade moving away from off-the-shelf chips toward custom silicon it designs in-house. The company’s first Full Self-Driving computer chip, introduced in production vehicles in 2019, replaced Nvidia hardware that Tesla had previously relied on. That move was followed by the Dojo supercomputer program, built around Tesla’s own D1 training chips, intended to accelerate the neural network training behind its self-driving software.
AI5 and AI6 represent the next step in that trajectory, extending Tesla’s chip ambitions from in-car inference toward a broader mix of robotics and data center workloads. The strategic logic is straightforward: a company that controls its own silicon can tune performance and cost to its exact needs rather than depending on general-purpose chips designed for a wider market. The tradeoff is that Tesla now needs manufacturing partners capable of producing cutting-edge chips at the volumes its vehicle and robotics businesses require, which puts it in direct, high-stakes negotiations with foundries like Samsung and TSMC.
That dependency is also why Tesla’s compute roadmap keeps surfacing in coverage of its other major bets, including the Cybercab and Megapack production plans that lean on the same chip and battery supply chains, and the ongoing debate over Tesla’s robotaxi fleet expansion, which depends heavily on the compute hardware generation these new chips are meant to deliver.
Samsung Foundry’s High-Stakes Bet on Tesla
For Samsung Foundry, landing Tesla as a customer for AI5 and AI6 carries significant strategic weight. Samsung has spent years trying to close the gap with TSMC in advanced-node manufacturing and has looked to marquee customers to validate its process technology and attract further business. A long-term relationship with Tesla, one of the most visible AI-chip customers outside the traditional hyperscalers, would reinforce that validation.
That strategic importance is precisely what makes the pricing allegation sensitive. If Samsung felt compelled to accept unusually aggressive terms to win or retain Tesla’s business, it raises questions about how much leverage large AI-chip customers now hold over foundries competing for a limited pool of marquee contracts. Samsung’s broader memory and foundry businesses are already under pressure, a dynamic visible in recent coverage of the HBM4 memory shortage squeezing Samsung and its rivals, which shows how stretched the company’s manufacturing capacity has become across multiple product lines simultaneously.
TSMC’s Position in the Triangle
TSMC remains the dominant force in contract chip manufacturing and a key alternative to Samsung for companies like Tesla. The unconfirmed claim that Tesla turned to Samsung partly because TSMC would not accept Tesla’s preferred terms is worth treating carefully, since it has not been verified by either company. But it fits a pattern that has played out across the AI chip industry over the past several years: as demand for advanced-node capacity has outstripped supply, foundries have gained more leverage to reject customer demands they consider unfavorable, rather than compete purely on price.
If that dynamic holds, it would mean Samsung, as the foundry with more capacity to fill relative to TSMC’s order backlog, may be more willing to accept demanding terms simply to secure volume. That theory remains speculative without confirmation from either Tesla or TSMC, but it offers a plausible explanation for why Tesla’s negotiating posture, as described in the allegations, differs depending on which foundry it is dealing with.
Why Custom AI Silicon Contracts Have Become This Contentious
Tesla is far from the only company pushing foundries and chip designers to the edge of what they’ll agree to. Across the industry, companies racing to build differentiated AI compute, from hyperscalers to automakers, are negotiating foundry agreements worth billions of dollars, and the competitive stakes make every contract term a point of friction. Microsoft’s push into custom silicon with its own accelerator hardware, detailed in coverage of the Maia 300 chip’s challenge to Nvidia, illustrates how many large tech companies are now following the same playbook Tesla pioneered: design your own silicon, then fight hard over who manufactures it and on what terms.
That broader context matters here. The same forces pushing up prices for AI accelerators generally, visible in reporting on how Nvidia’s B200 pricing has shifted amid major cloud deals, are also pressuring every company trying to build cheaper, custom alternatives. Tesla’s alleged hardball tactics with Samsung, whether or not the specific claims hold up, reflect how much is riding on getting AI5 and AI6 manufactured at a cost that supports Musk’s stated performance-per-dollar targets.
What’s Confirmed and What Remains Unverified
Given how quickly allegations like these can harden into accepted fact online, it is worth separating what the reporting actually supports from what remains speculative.
| Claim | Status |
|---|---|
| A Korean semiconductor engineer (@TTHUEONE) made the allegations publicly on X | Confirmed as reported |
| Tesla, Samsung, Samsung Foundry, TSMC, and Terafab are named in the claims | Confirmed as reported |
| Tesla allegedly sought lower AI5/AI6 prices plus internal fab info sharing and audit rights | Allegation, unconfirmed by Tesla or Samsung |
| Terafab allegedly recruited Samsung employees from the Taylor, Texas facility | Allegation, unconfirmed by Tesla or Samsung |
| Tesla picked Samsung Foundry because TSMC would not meet its demands | Unconfirmed claim |
| Specific prices, contract values, or employee headcounts | No verified figures published |
That gap between allegation and confirmation is exactly why outlets covering the story, including TrendForce, have been careful to frame the claims as allegations rather than established facts. Readers should expect the picture to shift, in either direction, as Tesla, Samsung, or additional reporting weighs in.
Market Impact: Why This Matters Beyond a Single Tweet
Even unverified, allegations like these can move markets and reputations because they touch a nerve that investors already watch closely: how durable are the supply relationships underpinning the AI buildout. Samsung’s foundry unit has staked significant capital on its Taylor, Texas expansion, and any suggestion that a marquee customer is extracting unusually favorable terms, or that staff are being poached mid-ramp, raises questions about the return on that investment.
For Tesla, the reputational risk cuts differently. The company has built part of its investment narrative around vertically integrated hardware, from battery cells to its own AI chips. Allegations that it strong-arms suppliers add to a pattern of scrutiny Tesla has faced over supplier relationships in other parts of its business, and they arrive at a moment when Tesla’s broader AI and compute ambitions are already under a microscope amid ongoing coverage of its robotaxi rollout and data center buildout plans.
For TSMC and other foundries watching from the sidelines, the dispute is a reminder that even the largest AI-chip customers can become difficult counterparties once a contract is signed, which may factor into how aggressively rival foundries bid for Tesla’s future chip generations beyond AI6.
How This Compares to Past Foundry Disputes
Tension between chip designers and the foundries that manufacture their silicon is not new. Apple’s shifting balance between Samsung and TSMC over the past decade, driven largely by yield and pricing disagreements, reshaped how both foundries compete for premium customers. Qualcomm has similarly split orders between foundries to maintain leverage. What distinguishes the current Tesla allegations is the addition of a workforce-poaching claim layered on top of the pricing dispute, combining a commercial disagreement with a talent-acquisition controversy in a single story.
That combination is likely to draw more attention than a pricing dispute alone would, because it touches two separate anxieties in the semiconductor industry at once: the fear that AI-chip customers are gaining outsized leverage over manufacturing partners, and the fear that talent, not just capacity, is the real scarce resource in the current AI buildout.
What Happens Next
Several outcomes are plausible from here, though none can be stated with certainty given how little has been officially confirmed.
- Tesla or Samsung could issue a formal statement addressing the allegations directly, which would either confirm, deny, or reframe the claims currently circulating.
- Additional reporting could surface more detail on the alleged contract terms, giving a clearer picture of whether Tesla’s requested terms were accepted, negotiated down, or rejected.
- Samsung could face pressure to clarify how many, if any, Taylor facility employees have left for Terafab, and under what circumstances.
- TSMC’s role in the story could be clarified, either confirming or dismissing the claim that it declined to meet Tesla’s terms for AI5 and AI6.
- The dispute could fade without further confirmation, joining a long list of unverified industry allegations that never receive an official response from the companies involved.
What seems likely regardless of how the specific allegations resolve is that scrutiny of Tesla’s foundry relationships will intensify as AI5 and AI6 move closer to production, simply because so much of Tesla’s future compute roadmap, from FSD to Optimus to its data center ambitions, depends on getting those chips built on schedule and at a defensible cost.
The Bigger Picture for AI Chip Supply Chains
Beyond the specifics of this one dispute, the episode is a useful snapshot of how strained the AI chip supply chain has become heading into 2027. Every major AI compute buyer, from cloud hyperscalers to automakers, is simultaneously trying to secure capacity, negotiate favorable pricing, and in some cases build out its own manufacturing capabilities, all while competing for the same limited pool of advanced-node fabrication slots and skilled engineering talent. That is the environment in which claims like these land with outsized weight: not because any single allegation is proven, but because they plausibly fit a supply chain already stretched thin by demand that keeps outpacing capacity.
Whether or not Tesla’s conduct matches the engineer’s description, the underlying pressure on foundries, memory makers, and the engineers who keep both running is real and well documented across the industry. That pressure is unlikely to ease soon, which means disputes like this one, confirmed or not, are likely to keep surfacing as more companies chase their own version of Tesla’s custom-silicon strategy.
Frequently Asked Questions
Who made the allegations against Tesla?
A South Korean semiconductor engineer using the X handle @TTHUEONE posted the claims publicly. The engineer’s identity beyond that handle has not been independently confirmed in the reporting reviewed for this story.
What exactly is Tesla accused of doing?
Tesla is accused of pushing Samsung Foundry for aggressively low prices on its AI5 and AI6 chips while also requesting internal fabrication information and audit rights as part of the manufacturing agreement. Separately, a Tesla-linked project called Terafab is accused of recruiting employees away from Samsung’s Taylor, Texas facility.
Have Tesla or Samsung responded to the claims?
As of this writing, neither company has publicly confirmed or denied the specific allegations. The claims remain unverified pending any official statement.
What are the AI5 and AI6 chips used for?
They are reported to be the next generations of Tesla’s custom silicon, intended to succeed the AI4 chip currently used in Tesla’s Full Self-Driving computer and to support broader workloads including robotics and data center inference. Specific technical specifications have not been officially confirmed.
What is Terafab?
Terafab is described in the reporting as a Tesla-linked fabrication project. It is named in the allegations as the entity that recruited employees away from Samsung’s Taylor, Texas semiconductor facility, though the scale and specifics of that recruitment have not been independently verified.
Why would Tesla ask for audit rights at a Samsung fab?
Large chip customers sometimes negotiate for information access or audit rights to verify yield, quality, and capacity commitments on major manufacturing contracts. The engineer’s allegation suggests Tesla’s requested scope went beyond what Samsung considered standard, but the exact contract language has not been made public.
Did Tesla pick Samsung because TSMC refused its terms?
That is an unconfirmed claim included in some of the reporting. Neither Tesla nor TSMC has verified or denied it, so it should be treated as speculative rather than established fact.
Could this affect Tesla’s chip production timeline?
There is no confirmed evidence that the dispute has affected production schedules for AI5 or AI6. Any impact on timelines would depend on how Tesla and Samsung resolve the underlying contract terms, which remains unknown publicly.