A long-form article republished by WuBlockchain says the layoffs spreading across Web3 are being explained in public as an AI story, but the pressure runs deeper. The piece argues that for many firms, especially crypto exchanges, the real issue is financial strain and a business model that no longer works as it once did.
The article opens with a broader tech backdrop. In the first half of 2026, the US technology sector cut nearly 140,000 jobs. Amazon reduced its workforce by 9%, and Meta cut 10%. Both cited AI as a central reason, saying the technology is changing everything and forcing companies to streamline. The report says more than 56% of layoff events in 2026 explicitly named AI, automation, or machine learning as a cause, and that AI has been the top stated reason for corporate layoffs in the US for four straight months. At the same time, nearly 60% of companies admitted they had packaged layoffs or hiring freezes as “AI-driven” even though the actual pressure came from finances.
According to the article, that pattern has not stopped in Silicon Valley. It has spilled into Web3, where technology and finance intersect and where the shock has been particularly sharp. Large-scale cuts have lasted for more than half a year. Since the start of this year, and especially in recent months, talk of staff reductions, team reshuffles, and employee movement around major trading platforms has become common on X, Reddit, Xiaohongshu, Maimai, and in private coffee chats among workers. The piece says BitMEX, once a major name, has largely faded from the mainstream view, while smaller platforms are exiting or shrinking business lines. In an environment where talent and attention are being pulled toward AI, layoffs in Web3 have come to look unavoidable to many people inside the sector.
People were cut before they understood they were on the list
One interview subject, Kevin, said he had only three days between receiving notice and his last day. He had previously worked at a major internet company, then moved into a leading crypto trading platform after being drawn by Web3 pay and the sector’s narrative. He later learned that the decision to remove him had actually been made more than a month earlier.
What stood out to him was the lack of visible warning. Work continued as normal. Meetings stayed on the calendar. Messages were answered. Then HR contacted him, without a reason he considered convincing and without a clear performance issue. Looking back, Kevin said the only possible signal was that two people had already left the 10-person group before him. At the time, the explanation inside the team was that they were “not a fit” or were leaving for “lighter work elsewhere.” Kevin told Zhangsheng BeatZ that, in hindsight, those departures may already have been part of a process to push people out.
Richard, who worked at a smaller exchange, described an even harsher method. After spending nine years as a full-time father, he returned to the workforce and found a job at a crypto exchange that was not especially large. He was soon laid off together with many colleagues. One morning, he opened his computer and found that his system access had already been disabled. He first assumed it was a technical problem. Then he checked the work chat and saw about 40 colleagues asking the same question: why could they no longer log in? Hours later, they received a cold layoff email in their personal inboxes. The action was immediate.
Richard said what made the experience worse was what had happened just before that. A manager had hinted that one of the developers reporting to him “might need adjustment.” Richard tried to help the employee stay. He even rearranged work assignments to show the person was hard to replace. Before he could formally submit that plan, both of them were out.
Another former exchange employee, Xiaoyu, gave Zhangsheng BeatZ a similar account. At her former company, the first sign of layoffs was often a batch shutdown of Slack accounts and email access. Whenever someone suddenly vanished from Slack, coworkers would rush into private chats and send over phone numbers and LinkedIn profiles while they still could. No one knew who would be next. Xiaoyu said that when her turn came, her manager sent a Slack message asking if she had time for a call. Before she could answer, all access had been revoked.
For some teams, the cuts came quarter after quarter
Kevin said the layoffs did not stop after he left. In the following months, his former group kept shrinking until only two people remained. At the platform where he worked, roughly 10% of staff were cut each quarter, adding up to about 40% over a year.
The article also points to Coinbase. In May, the company announced about 700 global job cuts, describing the move as an “AI-native reorganization,” or about 14% of staff. But the report says Zhangsheng BeatZ heard from people familiar with the matter that the impact in Coinbase’s India office was far greater. Former employees said about 90% of staff in India had left, across all departments, not only sales. Only a very small number of engineers regarded as top-tier were invited to relocate to Canada and continue working there.
The article says the main reasons behind the India cuts were high costs and the time difference with the US. According to the piece, Coinbase paid Indian SDE2 engineers around 7.5 million rupees, or about CAD 110,000, roughly in line with the pay level of mid-level engineers in Canada. It adds that at many high-paying product companies, Indian architects can earn even more than peers in the European Union.
The report says many exchanges have been accused of failing to reach agreement with employees over severance, then designating the same day as the worker’s last day and cutting off access immediately. It also mentions BitMart, saying the exchange, which has recently shut down, had already started eliminating entire departments in May.
Zhangsheng BeatZ says layoffs tend to cluster around specific dates for a reason. Around June 30 is described as one of the industry’s peak layoff periods because July brings fresh financial reporting for investors. Remove a batch of people, cut a layer of expenses, and the profit-and-loss statement improves at once. In that framing, layoffs are not only about cost reduction. They are also part of financial narrative management.
Severance pressure and efforts to make layoffs look like something else
The article says that, across Web3, very few layoffs end with compensation that employees consider fair. Several interview subjects described the speed with which exchanges sever contact and revoke access. Once internal systems are gone, so are colleague directories and many of the channels workers would need to defend their own interests.
According to a person cited by Zhangsheng BeatZ, operations and product staff who work in physical offices or overseas offices are more likely to receive standard handover periods and compensation. Many technical workers, however, operate remotely, and those are the people who can be dismissed almost instantly. The article ties that to a practical problem: many IT staff are based in mainland China, while the exchange itself is registered abroad. Even if workers want to challenge the process, cross-border enforcement is costly and difficult, so most people do not pursue it.
Even a short buffer period may not help. Kevin said HR asked him to fill in a reason for departure in the company system and tried to persuade him not to select “dismissed by company.” He said they warned that if he chose that option, future background checks might not go through and negative comments could be made about him. The pressure was to choose “personal reasons” instead. If the worker does that, the company does not need to pay extra compensation. Kevin said he received no severance. The company only paid salary and overtime up to his last day.
He also said there were signs he failed to read in time. His relationship with his direct manager had become less smooth, and he could tell the manager no longer liked him as much. Inside a fast-moving organization, though, subtle shifts like that are easy to ignore until the decision is final.
Surveillance software, KPI traps, and tests used as layoff tools
The article argues that during mass layoffs, exchanges often work hard to make layoffs appear to be ordinary performance management. Zhangsheng BeatZ says many interview subjects described a process in which major exchanges send company laptops to new hires before they start, with monitoring software installed inside. The software is said to track keystroke frequency and mouse-click behavior, and the resulting data can become part of performance evaluation.
The report says one exchange employee was fired the day after using a company-issued computer to watch an iQIYI drama for a short time. Another common practice, it says, is to assign KPIs that are nearly impossible to achieve and then use poor results as grounds for dismissal, citing “failed performance” or “failure to meet company requirements.” In this way, a layoff can be presented as a compliant personnel decision and severance obligations can be reduced.
The article adds that a former employee posted on X about a period when an exchange ran recurring “Web3 industry knowledge” tests and forced them into the KPI system. Failing the test could lead directly to dismissal.
The piece compares the layoff wave to a tornado: fast and violent. Yet because many workers have spent so long under high-pressure monitoring, few openly discuss what everyone can see.
Those who stayed were not necessarily in a better position
Xiaoyu said that after each round of layoffs, the so-called survivors often envied the people who had already left because at least their fate was decided. Those who remained lived like startled birds, never sure whether they would be next. She said morale had turned sharply negative once layoffs began, with a heavy sense of defeat settling over day-to-day work.
Richard described a similar change in atmosphere. Earlier, teams were still under intense pressure and product iterations were fast, but people were busy with real work: shipping updates and building features. During the layoff period, he said, the busyness changed. More of it was driven by management-imposed procedures. Assessment mechanisms were tightened, employees had to check in on the hour, and meetings became more frequent.
The article spends time on “stand-up meeting” culture at exchanges. Stand-ups are supposed to make meetings short and efficient. Richard said the tool had turned into a drain. His exchange held two stand-up meetings a day, yet no one knew where the product was actually heading. In half a year, the company went through three project managers. By the end, the product management team was almost empty. People still had projects in progress, but if a key person was fired in the middle of the day, or even minutes before a meeting, work could come to a stop immediately.
Richard also said the exchange used outsourced teams whose pay was noticeably higher than that of full-time staff. Later, after speaking face to face with two colleagues, he learned that an executive had held back employee raises for two years. His conclusion was blunt: management did not truly care about costs running out of control because what mattered more was power and control, not technology or product quality.
Kevin said he felt something similar. The exchange where he worked increasingly reminded him of an aging state-owned enterprise. At a time when the wider exchange sector was facing frequent security incidents, the technical team did not receive more support. Instead, it became more risk-averse. He said people no longer wanted to try anything bold. They just wanted to avoid mistakes, and the overall feel was that of a highly bureaucratic organization.
John, who grew up abroad, had already run out of patience with that kind of environment before being laid off. From the day he joined, he said, the company’s “Chinese culture” was overwhelmingly strong. Chat logs, JIRA, and meeting notes were almost entirely in Chinese. Foreign employees who did not speak Chinese well could feel excluded. The pace was intense, the atmosphere strict, and there was a performance review every quarter.
Because people were spread across time zones, logging in outside normal hours was routine. John said his weekly stand-up was scheduled for Sunday night, which regularly cut short his weekend plans. A QA colleague based in the US time zone often sent messages at 11 pm. He described the team as effectively “24 on call,” and said it was common to see colleagues committing code at 2 am on Saturday. In his words, there was no meaningful work-life balance, only work, life, and more work.
Internal power struggles and the logic of loyalty
Richard said he joined his company during its strongest period and watched the whole decline unfold from the inside. What struck him most was the struggle for power among senior management. In his view, it was more naked and more chaotic than ordinary office politics.
He said the company’s partners fell into a severe trust crisis because of government investigations and potential lawsuits. One side’s CTO and CFO believed they had been misled by the other side or had not been properly supported when regulatory trouble appeared. In the end, the partners split. One side took a core team and a senior employee, formed a “board,” and created a new company that became the de facto developer of the old product. Partners who had once called each other friends turned into counterparties within a month. By February, the new company was pushing ahead at a rate of two new products a week. All of that happened around the time Richard resigned.
At the lower levels, employees had neither the right to know nor the ability to choose. The article says they became casualties of internal conflict and instability. It goes further and argues that in Web3, many project founders and even exchange CEOs are merely public-facing figures. The real decision-makers are often behind the scenes, and the skill needed most in visible leadership positions is not always innovation or technical ability, but loyalty.
Kevin called the culture toxic from the top down. In his view, the people who survive in that system tend to become exactly that kind of person. If someone climbs upward, he said, the environment often reshapes them. The people who get promoted are usually skilled at political maneuvering, strong at managing upward, and forceful with subordinates.
Those who are not part of the inner circle can be eased out step by step. First they stop being invited to meetings. Then key decisions are made without them. Next they are shifted to marginal roles, away from the core business. Then weekly reports are no longer required and new assignments stop arriving. By the time a replacement is already lined up, the person realizes they have effectively been hollowed out. Kevin said Glassdoor reviews often note that colleagues themselves are supportive and decent, yet the overall structure feels like a palace court where even wording in front of superiors has to be carefully managed.
The exchange revenue engine is weakening
The article then turns to a structural question. Kevin’s view is direct: “the whole crypto business model has collapsed.”
As the piece lays it out, exchanges historically depended on two core revenue streams: trading fees and listing fees. When the market was hot, new projects flooded in, retail users traded heavily, fee income rose, and teams expanded with it. Now, Kevin says, many of the projects that come to list have already shown what they are after: they want to make money and leave.
Listing fees, he argues, have become a serious problem of their own. Exchanges charge very high amounts to projects. A small project may have to pay hundreds of thousands of dollars just to get listed, while the token’s market value after launch may be only in the tens of millions. Kevin’s description is that exchanges have drained the ecosystem dry. Starting a crypto venture is too expensive, and retail buyers are no longer willing to absorb the cost. He sees a downward spiral: weaker project quality, more tokens breaking below issue price, retail participants leaving, volumes shrinking, fee income falling, and listing fees being pushed up again to compensate.
The rise of onchain derivatives platforms led by Hyperliquid adds another layer of pressure. The most profitable part of a centralized exchange, derivatives trading, no longer has to happen within a centralized system. That means exchanges are now losing traffic and economics in one of their most lucrative segments.
Liquidity stress, the October 10 liquidation shock, and cautious venture capital
Broader market conditions are making the slide worse. Multiple interview subjects pointed to the industry-wide liquidation event on October 10 last year. According to the article, every open leveraged position above 2x was forcibly liquidated that day. Retail investors were hit hard and have not recovered, and confidence across the industry took lasting damage.
John told Zhangsheng BeatZ that a number of mid-sized Web3 institutions managing between $100 million and $500 million are shutting down. Old fundraising approaches and DeFi yield strategies are becoming harder to sustain. Since last summer, he said, the drying up of crypto liquidity has been “very severe.” Almost all altcoins launched in early 2025 are approaching zero, with extremely low book value. Over-the-counter volume is weak. In market making, he said, there is little worth doing outside RWA-related business.
He also mentioned a friend running a crypto-neutral strategy at a market maker. The friend improved market share and profit per trade through strategy upgrades, yet the firm’s total profit still fell sharply. Profitability across the business dropped to around 30% of previous levels, and the friend eventually lost the job in a cost-cutting round.
The pullback is not limited to market makers and quant firms. Kevin said Web3 venture capital has become very cautious in both check size and deal count, to the point of being functionally inactive. Even when investors do write checks, the amounts are far smaller than before. In his words, funding in this cycle is down 80%, and almost nobody wants to invest in crypto. That, he said, means the next bull market may not have many credible new listings for retail users.
Projects themselves are also struggling. Kevin’s assessment is that apart from some teams with Web2 revenue on the B2B side, most projects have neither meaningful B2B income nor durable consumer-side revenue.
Many leave for AI, but the move is not clean or easy
The report says the challenge for people cut from exchanges is not simply whether to switch sectors, but whether there is anywhere strong to go. Kevin left his exchange job for an AI startup. Zhangsheng BeatZ says he is far from alone. Most people leaving Web3 are heading into AI. The logic is straightforward: AI is the hottest track right now, financing is active, jobs are more available, and crypto and AI share some practical overlap in growth, user acquisition, and global operations.
Richard’s disillusionment with Web3 is more complete. He said the exchange where he worked was full of incompetent people from top to bottom, from partner infighting to disengaged staff lower down. He also moved into AI and left crypto behind.
By contrast, only a small number of people manage to move into traditional sectors. The article says some technically strong specialists in trading systems and risk control have joined traditional market makers and quant firms. A few people in operations, BD, and compliance have entered conventional brokerage systems as Hong Kong and US equity markets stay active. But these are exceptions. More often, people cut by one exchange end up at a smaller exchange in the next tier.
Bias against Web3 resumes runs deep
The report says traditional industries can be more hostile to a Web3 background than many insiders expect. Zhangsheng BeatZ says some HR departments in traditional finance have indicated that they may reject candidates outright if they see current employment at a Web3 company on the resume.
In the eyes of many traditional finance professionals, the article says, crypto resembles a giant “cockroach motel”: a place associated with regulatory gray areas, speculative culture, and results that are hard to verify. People coming from that world are viewed as carrying an original stain.
The bias is not limited to finance. The article says serious AI companies working on large models and infrastructure can also be skeptical of candidates from Web3. In their view, “growth” in crypto has often been built on speculation and narrative rather than real technical barriers. An operations employee from an exchange and one from ByteDance are not seen as equivalent by AI recruiters. The article argues that this may prove to be the most lasting cost for laid-off Web3 workers.
A colder winter than previous cycles?
The article closes by asking whether this downturn is different from the old cycle story. Competition around retail capital has changed. Prediction markets such as Polymarket and Kalshi, along with retail brokerage trading, are now competing for the same users and the same money. US retail investors are putting money into AI stocks and prediction markets rather than returning to crypto.
Some industry participants quoted in the piece say conditions are even worse than the crypto winter of 2022. Back then, at least retail traders were still in the market. In their view, the large-scale liquidation event on October 10 drove out the final group of leveraged retail players.
Even in a shrinking market, the article says, exchanges have not stopped fighting one another in destructive ways. According to a person familiar with the matter cited by Zhangsheng BeatZ, one exchange’s HR department even made “poaching talent from competitors with high salaries” part of its KPI. Workers would be lured over, then let go a few months later under various pretexts. The goal, the source said, was to disrupt rivals, gather intelligence, and gain customer resources, while the people being hired served as disposable tools.
The article compares that behavior to the food-delivery subsidy wars in China’s internet sector years ago. It cites a figure of more than RMB 220 billion, or about $31 billion, spent by Alibaba, Meituan, and JD.com on delivery subsidies over two quarters, a sum it says was close to what all companies worldwide spent on generative AI over a full year.
In the author’s framing, crypto exchanges are now replaying that script. The pie is shrinking. Retail users are leaving. Volumes are falling. Yet platforms are still poaching each other’s staff, undermining one another, and fighting costly battles for leftovers.
The piece returns to its original point in the end. For a long time, mass layoffs at trading platforms were explained away as part of the Web3 cycle or as fallout from AI. But if more than half of tech layoff events in 2026 named AI while nearly 60% of companies privately acknowledged finance as the real cause, then Web3 may not be any different.
The article lists the practices it sees as central to the problem: listing fees of hundreds of thousands of dollars charged to projects, large numbers of weak tokens pushed into the market and sold to retail investors who repeatedly lose money when prices break below issue levels, opaque performance systems and surveillance that erode trust and creativity, and management teams that spend scarce resources poaching from rivals instead of finding a new business model for a harder market. It ends not with an answer, but with a question: if today’s Web3 winter is not only a cycle, then who is responsible for the industry’s decline?
The piece says it is based on interviews with Kevin, Richard, Xiaoyu, John, and others. Their real names, employers, roles, and tenure lengths were withheld for anonymity.