Author: Jialiu, Zhangsheng BeatZ
'AI is the main reason for our layoffs.' This is almost the primary explanation given by all companies conducting layoffs today.
In the first half of 2026, the U.S. tech industry laid off nearly 140,000 people. Amazon cut 9% of its workforce, and Meta cut 10%. Their reasons for the layoffs were almost identical: AI is changing everything, and the company must streamline.
In fact, over 56% of layoff events in 2026 explicitly cited AI, automation, or machine learning as the reason. AI has been the top reason for corporate layoffs in the U.S. for four consecutive months. Ironically, nearly 60% of companies admit that they package layoffs or hiring freezes as 'AI-driven' when the real reason is financial pressure.
The shockwave of AI is not confined to Silicon Valley; it is reshaping the employment structure in almost all industries. As the intersection of technology and finance, the Web 3.0 industry is experiencing particularly intense impact. Large-scale layoffs in the Web 3.0 industry have persisted for over half a year, and they have been exceptionally fierce and rapid.
Since the beginning of this year, especially in recent months, news surrounding the layoffs, team restructurings, and personnel movements at leading trading platforms has densely appeared on X, Reddit, Xiaohongshu, Maimai, and in coffee chats among practitioners. The once-dominant BitMEX has almost faded from the mainstream spotlight, and smaller platforms are exiting or shrinking their business lines. With talent and attention being drained by the AI industry, layoffs in the Web 3.0 industry now seem inevitable.
The Sword of Damocles Has Finally Fallen
Kevin received his termination notice just three days before his last day.
Kevin had worked at a major internet company for several years before being attracted by the high salaries and narratives of the Web 3.0 industry, joining a leading trading platform. He later found out that his termination had actually been decided over a month prior.
During that period, he had hardly sensed any signals of impending layoffs. All work proceeded normally, meetings continued, messages were replied to. Until HR approached him—there was no reasonable explanation, no claim of poor performance—but the sword of Damocles still fell on Kevin.
In hindsight, the only signal might have been that two out of their original ten-person team had left before him. At the time, the explanation given was 'not a good fit' or 'found other, easier jobs.' 'Looking back now, they were probably already being pushed out then,' Kevin told Zhangsheng BeatZ.
The layoff method at Richard's small trading platform was even more extreme. After being a stay-at-home dad for nine years, he returned to the workforce and found a job at a small crypto trading platform. But he and many colleagues were soon laid off.
According to him, one morning he opened his computer as usual to start work and found his system access had been revoked. He initially thought it was a technical glitch until he opened the work chat and saw about forty colleagues asking the same question: 'Why can't I log in either?' No one knew what was happening. Panic spread through the group chat like water. Several hours later, they received a cold termination notice in their personal emails, effective immediately.
What chilled Richard even more was another incident. Shortly before being laid off, his manager had hinted that one of the developers under him 'might need adjustment.' Richard was still trying to help this colleague stay, even rearranging work assignments to prove the person was irreplaceable. Before he could submit the proposal, both of them ended up being laid off.
Another former employee, Xiao Yu, who previously worked at a crypto trading platform, described a similar layoff scenario to Zhangsheng BeatZ. At her former company, the first step of layoffs was to batch-disable employees' Slack accounts and cut off email access. 'Whenever we saw someone suddenly disappear from Slack, we would immediately rush into the private chat channel, scrambling to send our phone numbers and LinkedIn links,' Xiao Yu said. 'Because we didn't know who would be next, everyone wanted to stay connected while they still could.'
'When the layoff finally hit me, my manager messaged me on Slack asking if I had time for a call,' Xiao Yu said. 'Before I could reply, all my permissions were revoked.'
Layoffs Like a Tornado
Kevin revealed that in the months after he left, his team continued to lay off people, now down to just two members. His former trading platform laid off about 10% of its staff each quarter, accumulating to 40% over the year.
Coinbase announced global layoffs of about 700 people in May, officially defining it as an 'AI-native reorganization,' roughly 14%. However, according to information obtained by Zhangsheng BeatZ from informed sources, the impact on Coinbase's India office far exceeded that number. A former employee claimed that about 90% of the India office staff left, involving all business units, not just sales. Only a very few, considered top-tier engineers, were invited to relocate to Canada to continue working.
It is said that the main reasons for the large-scale layoffs at the India office were high costs and significant time differences with the U.S. Coinbase paid Indian SDE2 (mid-level engineers) salaries of about 7.5 million rupees, equivalent to approximately 110,000 Canadian dollars, on par with mid-level engineer salaries in Canada itself. In most high-paying product companies, Indian architects' salaries are even higher than their EU counterparts.
Many trading platforms have been exposed for employees failing to negotiate severance packages with HR, being told the same day would be their last workday, and having their system access cut off. The recently closed platform BitMart had entire departments laid off starting in May.
Furthermore, many platforms choose to conduct concentrated layoffs at specific times, not by coincidence. According to Zhangsheng BeatZ, a peak for industry layoffs occurred around June 30th. The reason is simple: new financial statements are due in July, and these figures are to be shown to investors. Laying off a batch of people and compressing expenses immediately makes the P&L statement look better. For trading platform management, layoffs are not just about cost reduction; they are also a form of financial narrative management. In front of investors, a streamlined statement is more persuasive than any explanation.
It's not just Kevin's and Richard's trading platforms; almost the entire Web 3.0 industry is undergoing large-scale layoffs, with only a minority receiving reasonable and satisfactory severance packages.
The situations encountered by the interviewees mentioned earlier are largely similar; trading platforms cut off contact and permissions very quickly during layoffs: 'Everyone's contact information is on there. Without these permissions, we don't even have a channel to fight for our rights.'
According to information obtained by Zhangsheng BeatZ from informed sources, operations and product roles, because they work offline or in overseas offices, still get normal handover time and compensation during layoffs. 'But many technical staff are remote, so they just fire them, quickly fire them. It doesn't really affect the company.'
Because many IT employees are based in mainland China, while the trading platforms are registered overseas. 'You're not physically there; the cost for an individual to pursue action is very high. It's just that amount of money, not affecting daily life, so most people don't want to and can't afford to make a fuss.'
Even with a few days of buffer, employees faced difficult situations. When communicating about the offboarding process, HR asked Kevin to fill out the reason for leaving in the system and advised him not to select 'company termination.'
'They would say, if you choose company termination, background checks won't pass you, they'll say bad things about you. So they pressure you to choose personal reasons for leaving.' Choosing personal reasons means the company doesn't need to pay any additional compensation.
Kevin ultimately did not receive any severance pay; the company only settled his salary and overtime pay up to his last workday. Reflecting later, Kevin himself also felt there were signals he hadn't read in the time before being laid off. For example, the working磨合 with his direct supervisor started to become less smooth, and he could clearly sense the other party didn't like him as much. But in an organization running at high speed every day, these subtle changes were easily overlooked until the moment the shoe dropped.
During the period of mass layoffs, major trading platforms are trying every means to create fancy names to make layoffs look like something else.
For instance, Zhangsheng BeatZ also learned from many interviewees: before employees join, leading trading platforms send out pre-configured computers. These computers have strict monitoring systems installed internally, capable of tracking employees' keyboard input frequency and mouse click behavior. This data is incorporated into performance evaluations.
It's said that a trading platform employee was fired the next day for using the company-issued computer to watch a show on iQiyi for a while.
There are also many common tactics, such as first setting almost impossible-to-achieve KPIs for employees, then after the evaluation period, terminating them citing 'poor performance' or 'not meeting company requirements.' Through this method, layoffs are packaged as compliant performance-based elimination, and the company avoids paying extra compensation.
A former trading platform employee revealed on X that during a layoff period, a trading platform held regular 'Web 3.0 industry knowledge' tests, making them mandatory for KPI assessments. Employees who failed the exams faced the risk of direct dismissal.
This massive wave of layoffs, like a tornado, came swiftly and fiercely. But because of the long-term high-pressure monitoring environment, everyone silently avoids mentioning the elephant in the room.
Under the Wind, A Chilling Silence
Compared to those who were laid off cleanly, those who remain are not necessarily luckier.
Xiao Yu said after each round of layoffs, survivors actually envy the colleagues who have already left because at least their fate is sealed. Those who remain live every day like startled birds; no one knows if they are next. Since the layoffs began, she could clearly feel the work mood become extremely negative, permeated with an unspeakable gloom, and no one had any drive for anything.
Richard also mentioned the subtle change in the work atmosphere during the layoff period. Previously, the work pace was tight, intense, and product iterations were more aggressive, but most of the time people were busy with real work, product updates, and feature iterations. The busyness now is completely different, mostly to satisfy management's fancy new requirements. Companies increased assessment mechanisms, required punctual check-ins, and meeting frequency became higher than before.
The 'stand-up meeting' culture at trading platforms was pushed to extremes during the layoff period. The original intent of stand-ups is to have quick meetings; standing is uncomfortable, so people get to the point. But according to Richard's description, at his trading platform, this tool meant to increase efficiency turned into a drain: two stand-up meetings a day, but even so, no one knew which direction the product was heading.
Three project managers were replaced in half a year, and the product management team was eventually almost empty. Many people had ongoing projects, but because key personnel were suddenly fired, sometimes even minutes before a meeting, this work had to stop abruptly.
According to Richard's description, there were even outsourcing teams at his trading platform, and these contractors' salaries were significantly higher than those of full-time employees. It wasn't until Richard later spoke face-to-face with two colleagues that he learned an executive had withheld salary increases for employees for two years.
Richard believes management didn't care about cost control because what they truly cared about was not technology or product, but power and control.
Kevin's feelings were the same. He increasingly felt his trading platform resembled a sluggish, bureaucratic state-owned enterprise. Against the backdrop of frequent security incidents across the crypto trading industry, the platform's tech team not only didn't get more resources but became a state of startled birds: not seeking merit, just avoiding blame.
'No one dares to take risks anymore. They just want their own work to not go wrong. The whole place feels like a state-owned enterprise,' Kevin said.
Before being laid off, John, who grew up abroad, had also long lost patience with such a work environment.
From the time he joined, he clearly felt the company had a particularly heavy 'Chinese culture.' Chat records, JIRA, meeting minutes—almost everything was in Chinese. Foreign employees who weren't good at Chinese felt excluded. The work atmosphere was extremely strict, the pace was fast, and there was a performance review every quarter.
Since people were distributed across different time zones, being online at unusual hours was commonplace. John mentioned that his team's weekly stand-up was scheduled for Sunday night. 'My weekend plans always ended early.' His QA testing colleague was in a U.S. time zone and often sent messages around 11 p.m.
'We always had to be on call 24/7,' John said, often seeing colleagues commit code at 2 a.m. on Saturday. 'There was no work-life balance here. The life rhythm was more like work, life, then more work.'
Deep Palace Politics, Favorites and Pawns
Richard joined during the company's peak and witnessed its decline from prosperity. What made him most唏嘘 was the 'power struggles' among the trading platform's management, which were more赤裸 and chaotic than other office politics.
The partners at his company developed a severe trust crisis due to government investigations and potential lawsuits. One side's CTO/CFO felt deceived by the other partner, or believed they didn't receive proper support when facing government issues. Eventually, the partners went their separate ways, announcing a split.
So one side took the core team and a senior employee to form a 'board,' establishing a new company that became the actual development party for the old product. Those once-called friends and partners turned into client relationships within a month. By February, the new company was pushing business at a rate of two new products per week. All this happened around the time Richard resigned.
Grassroots employees had neither the right to know nor the right to choose in these high-level power struggles. They were all casualties of internal strife and turbulence.
In the Web 3.0 industry, many project founders and even trading platform CEOs are just frontmen. This is an open secret within the industry, tacitly understood by almost all practitioners. The real decision-makers often lurk behind the scenes, and the foremost quality needed by those on stage is not innovation or technology, but loyalty.
'Toxic culture is transmitted from the top down. People who survive in this system are probably such characters. If you can climb up, you will inevitably be assimilated by this environment. If you are not such a person, you won't get promoted,' Kevin analyzed. 'Those promoted are almost always people skilled in political maneuvering, good at upward management, and assertive with subordinates.'
Those deemed not part of the 'inner circle' are cleared out by upper management using various methods. First, they stop inviting them to meetings, bypassing them for key decisions. Then they transfer them to marginal positions, away from core business. Next, they no longer require weekly reports and stop assigning new tasks. By the time replacements are already arranged, they finally realize they have been sidelined.
'So the entire system is very toxic,' Kevin said. 'You can check Glassdoor; people generally think colleagues are nice, willing to support each other, and have good personalities. But the whole system is like a deep palace. You can't say the wrong thing in front of superiors and must be careful with your rhetoric.'
When the Nest Falls, No Egg Remains Intact
'I think the entire Crypto business model has collapsed,' Kevin said.
In the past, trading platforms' core revenue relied on two things: trading fees and listing fees. When the market was hot, new projects flooded in, retail investors swarmed to trade, fees and listing fees soared, and teams expanded accordingly. 'But now all the projects seeking listings have been proven; they all just want to make money and then leave.'
The issue with listing fees is equally severe. According to Kevin's description, trading platforms charge project teams extremely high fees; a small project might have to pay hundreds of thousands of dollars just for listing, while the post-listing market cap might only be tens of millions. 'The trading platforms are eating the entire ecosystem to extinction. On one hand, the cost to start a business in crypto is too high; on the other hand, retail investors are no longer buying.' In his view, this is a downward spiral: project quality declines, more projects open below issue price (破发), retail investors leave, trading volume shrinks, fee income decreases, and listing fees are forced to rise.
The rise of on-chain derivatives platforms like Hyperliquid has made the situation for centralized exchanges even more passive. The most profitable segment for centralized exchanges—derivatives trading—is no longer confined to happening within their own systems.
Market-level shocks are also accelerating this spiral.
Many interviewees mentioned the large-scale industry-wide liquidation event on October 10th last year, whose negative impact has dealt a profound blow to the confidence of all practitioners. All leveraged positions with leverage exceeding 2x were forcibly liquidated that day, retail investors were wiped out, and have not recovered since.
When the nest falls, no egg remains intact; no one is spared. The困境 of trading platforms ripples out to the entire industry.
John told Zhangsheng BeatZ that many mid-sized Web 3.0 institutions with assets under management between $100 million and $500 million are shutting down. Old financing strategies and DeFi yield strategies are becoming increasingly difficult to sustain. And since last summer, the liquidity drought in cryptocurrencies has been 'very severe.' Basically, all altcoins launched around early 2025 are trending towards zero, with extremely low book values. Over-the-counter (OTC) trading volume is dismal, and besides RWA-related business, there's almost nothing worth doing in market making. A friend of John's working in crypto-neutral strategies at a market maker told him that even though they improved strategies to increase market share and profit per trade, the company's total profits still dropped significantly, generally shrinking to about 30% of original levels. John's friend was eventually laid off due to cost-cutting.
It's not just market makers and quant firms. Kevin mentioned in the interview that VC investments in the Web 3.0 industry are currently very cautious in terms of amount and number of deals, basically in a state of not investing. Even when they do invest, the amounts are significantly reduced. 'VC investment volume has shrunk by 80% this cycle. Hardly anyone is investing in crypto anymore. So when the next bull market comes, there won't be many good projects for retail investors to list either.'
The situation for project teams is equally tough. Kevin's assessment is: 'Except for some projects with Web 2.0 revenue on the B2B side, the vast majority of projects have neither B2B nor B2C revenue.'
Crypto is Like a Roach Motel
A fine bird chooses a tree to perch on, but for those laid off from crypto trading platforms, the problem isn't choosing the right tree, but whether there are any trees left to choose.
After leaving the trading platform, Kevin joined an AI-related startup. He's not an isolated case. According to Zhangsheng BeatZ, the vast majority of practitioners leaving the Web 3.0 industry have flocked to the AI industry. This is not hard to understand: AI is the hottest sector currently, with active financing, ample positions, and many skills directly transferable due to commonalities in channels and attributes between crypto and AI, both emphasizing growth, user acquisition, and global operations.
Richard's disappointment with the Web 3.0 industry is even more profound. In his view, the trading platform he worked at was filled with incompetent people from top to bottom, from partner infighting to基层 employee slacking. 'Even today, people in the cryptocurrency circle are still a bunch of self-righteous, arrogant people.' He later also turned to the AI industry, completely leaving the crypto world.
In contrast, not many can truly transition into traditional industries. A small portion of technically solid trading system and risk control talent entered traditional market makers and quantitative firms. Some operations, BD, and compliance personnel, taking advantage of active Hong Kong and U.S. stock markets, transitioned into the traditional brokerage system, but these are minorities. The outcome for more laid-off trading platform employees is flowing to the next tier of smaller platforms.
Because the discrimination against the Web 3.0 industry from traditional sectors is deeper than many imagine.
Zhangsheng BeatZ learned from some traditional financial HR departments that during recruitment, they directly淘汰 candidates whose resumes still show employment at Web 3.0 companies. In the stereotypical impression of many traditional finance practitioners, the crypto industry is like a giant 'roach motel,' implying regulatory gray areas, speculative culture, and unverifiable performance. People coming from here naturally carry original sin in their eyes.
Even within the AI industry, similar prejudice exists. Some serious AI companies focused on large models and infrastructure also have reservations about candidates with Web 3.0 backgrounds. In their view, the 'growth' in Web 3.0 is built more on speculation and narratives than real technological barriers. An operations person from a trading platform versus an operations person from ByteDance have vastly different含金量 in the eyes of an AI company's HR.
This might be the most profound cost experienced by Web 3.0 practitioners who have undergone layoffs.
This Winter is Colder Than Before
Every industry has cycles. But this round of winter for the Web 3.0 industry might be different from before.
Compared to the past, the competitive landscape in the cryptocurrency field has fundamentally changed. Prediction markets like Prediction Market, Polymarket, and Kalshi, along with retail brokerage trading, are all competing for the same pool of funds from the same batch of retail investors. U.S. retail investor money is flowing into AI stocks and prediction markets, not back into crypto.
Some practitioners even believe the current situation is worse than the crypto winter of 2022. In 2022, at least there were retail investors present. Now, the October 10th massive liquidation washed away the last leveraged retail investors.
To see if an industry is youthful or aged, don't just look at its revenue; look at what it fights for.
Even in such a 'volume-shrinking' market, the struggles and dirty tricks among trading platforms haven't stopped. According to information obtained by Zhangsheng BeatZ from informed sources, the HR department of one trading platform even listed 'poaching people from competitors with high salaries' as a KPI metric. They would hire them for a few months, then fire them under various pretexts, disrupting competitors' team节奏, obtaining intelligence and client resources. The poached individuals were just disposable tools.
This reminds the author of the internet industry's food delivery wars a few years ago, where the smartest people spent hundreds of billions in profits互相消耗. Alibaba, Meituan, and JD.com—three Chinese internet giants—burned over 220 billion RMB, approximately $310 billion USD, on food delivery subsidies in two quarters,接近 the total global corporate spending on generative AI for an entire year.
Today's crypto trading platforms are replaying the same script. The entire industry's pie is shrinking, retail investors are leaving, trading volumes are萎缩, but platforms are still poaching talent, publicly bad-mouthing each other, and engaging in attrition wars over the existing market.
In the past, we often attributed massive trading platform layoffs to the cyclical nature of the Web 3.0 industry and the impact of AI. As mentioned at the beginning of the article, over half of tech layoff events in 2026 used AI as a reason, but nearly 60% of companies admitted the real cause was financial pressure.
The Web 3.0 industry is no exception.
Charging project teams hundreds of thousands in listing fees; listing大量 quality-poor tokens, causing retail investors to lose repeatedly in破发; consuming employee trust and creativity with opaque performance reviews and monitoring systems; in winter, instead of thinking about new business models, spending resources on poaching talent from rivals.
If today's Web 3.0 industry winter is not the destiny of a cycle. Then whose fault is the decline of the Web 3.0 industry?
