A note to the reader: For those unfamiliar with the Polis Labs story, the founders met while attending Network School’s inaugural cohort in 2024. We have nothing but respect for Balaji, who has been willing to put an incredible amount of skin in the game to prove the validity of his vision. We wish him and all current and future Network Schoolers the best of success in their future endeavors. That being said, all failed projects require a post-mortem. And since assessing what works and what doesn’t for similar communities is the key mission of Polis Labs, what follows is our diagnosis of what went wrong in Malaysia and what we can learn from it.
In what appears to be a textbook application of punitive bureaucracy, Balaji’s Network School project in Forest City, Johor, Malaysia, had its business license revoked by the Iskandar Puteri City Council and was ordered to cease all operations last week. The council’s shutdown order followed a dodgy series of punishment-is-the-process-type events, including an intrusive inspection of the immigration documents of all 200+ Network School residents by Malaysian authorities. This series of events was apparently sparked by a shadowy activist group that accused Balaji of breaking Malaysia’s immigration laws by accepting Israeli passport holders into Network School (Malaysia does not formally recognize the state of Israel and bars most of its passport holders from the country). Although this was proven false, the government scrutiny continued. Network School was eventually cited for minor regulatory infractions* that would ordinarily merit a warning letter or a small fine, but instead led to the above-mentioned business license revocation.
*Something clearly went wrong with the relationship among Balaji, Network School, and the Johor state authorities. We’ll probably never know what it was. But this type of over-the-top technical enforcement is almost always a visible manifestation of an opaque government decision-making process that obscures the actual issue behind layers of red tape.
As of today, it does not appear that Balaji intends to appeal the Iskandar Puteri City Council's decision; instead, he has petitioned Malaysia’s Prime Minister directly for redress. However, hearing nothing but crickets from Kuala Lumpur, Balaji appears to have decided to pop smoke and search for greener pastures in Central Asia.
In a blindingly fast turnaround, he signed a five-year MoU with Kazakh authorities on 21 July. The final location isn’t fully nailed down, but according to regional reporting, the plan being floated would put Network School in Kazakhstan’s Akmola region, as part of a jointly-developed “Network City” in the Burabay resort area (a three-hour drive north of Astana), with a later expansion toward the planned “smart city” of Alatau, outside Almaty.
Mistakes Made All Around
Balaji’s a smart guy surrounded by smart people, so I expect him and Network School to survive this kerfuffle with Malaysia. But will they thrive in Kazakhstan? This is a bigger question, because mistakes were made with the Malaysian model.
Did the Johor state government and Malaysia in general bungle this excellent opportunity to attract talent and investment? Absolutely. Did Balaji and Network School follow the rules and comply with all Malaysian regulations? Sure looks that way. But, while I hate to say this, Balaji and Network School are victims of their own planning shortfalls, not bad luck. And if they don’t recognize that fact, then they face another expensive mistake in Kazakhstan. Chalking up the Malaysia episode to simply being unlucky means their model doesn’t need tweaking. But it does. And here’s why.
Predictably Unpredictable
There’s a raging debate among startup societies about which jurisdiction to choose as a physical base. Do you choose a locality with higher barriers to entry (higher cost of living, tighter visa requirements, stricter regulatory environment, etc.), or one that provides more operational flexibility but considerable regulatory uncertainty?
If you choose option B, you are opting for predictable levels of unpredictability and must plan accordingly for a much wider array of potential risks. It is no secret that (aside from Singapore) foreign businesses run a range of political, regulatory, legal, and financial risks when operating in Southeast Asia. These risks are not immeasurable. Political risk analysis, stakeholder mapping, and cultural sensitivity assessments are mature disciplines with decades of method behind them and are standard practice in extractives, infrastructure, and humanitarian work, among other industries. With the right tools and planning in place, predicting the unpredictable simply becomes part of the operating process. AI-driven automation makes the risk management process of identifying, assessing, and treating risks even easier to integrate into an organization’s decision-making.
Ultimately, it appears that Balaji based Network School in Malaysia because Forest City’s emptiness offered a great opportunity to grow, while the location was close to Singapore but without Singapore prices — a perfectly rational business decision. But the additional pitfalls and risks of working in Malaysia’s unpredictable jurisdiction, compared with Singapore’s more stable one, require extensive risk management planning to achieve long-term success; I’m not sure Balaji recognized this until it was too late in the game. But he will need to adjust before the venture in Kazakhstan, as he will find the same challenges in Central Asia as in Southeast Asia (and almost certainly more).
Protecting Local Buy-In
As part of the risk management process, we always recommend prioritizing relations with local communities. One of our key focus areas at Polis Labs is to protect local buy-in: "Network states, charter cities, and intentional communities almost always land somewhere. That somewhere already has people in it. A parallel society that imposes itself on the locals it arrived among has reproduced the failure mode it claimed to escape. We help founders build the relationships, agreements, and shared institutions that make the people already living in the territory genuine participants in the governance, not subjects of it.” Over the long term, this also prevents the development of colonial attitudes and the enactment of coercive measures against local and nonlocal populations. Polis Labs prides itself on providing advisory and preventative measures to steer longterm communities away from activities that embrace failed legacy governance structures.
With that throat-clearing aside, below are a few of our “iron rules” for relations with local communities.
Iron Rule 1: You will not stay a guest
In the post below, we examined the four categories of visitor: expat, immigrant, missionary, and guest. A key point we made was that the host community of any startup society gets a hand in determining this category, based on how the newcomers actually behave once they show up. One thing we didn’t note in that previous post is that a visitor’s status is also affected by the time horizon.
Expat, Immigrant, Missionary, Guest
The Polis Labs team was fortunate enough to attend the Infinita Crypto Cities Summit at Próspera earlier this month. Throughout the week, we had an opportunity to discuss "startup societies" encompassing a diverse range of initiatives, from network states and pop-up cities to online communities and charter cities. These endeavors seek to experiment with…
This time horizon will dictate how the new community is perceived locally, and for the newcomers, managing local perceptions should be a core planning function.
Because in the eyes of the locals, any guest becomes a resident within a particular timeframe. And a resident with a gated compound closed off to most locals becomes an expatriate enclave. And an enclave whose founder has written a book about detaching technology from national sovereignty, and once called that strategy “tech Zionism,” becomes a missionary to outside observers, whatever he or anyone inside the walls believes.
For a startup society to last, it must plan in year zero how it wants to be perceived by the local community in year five and build towards that outcome continuously. Language, local hiring, local capital participation, and civic presence are Day 1 agenda items that need a budget and a named owner.
A major potential failure we have observed in this movement is a mismatch in scale. You are not going to gain local buy-in by pairing what will be perceived as immigrant-scale commitments (buildings, five-year agreements, permanent campuses) with guest-scale investment in the host relationship. Locals watch leases get signed, foreign money move in, and other marks of a permanent arrival, often before anyone from the project has bothered to introduce themselves or explain their intended time horizon.
Iron Rule 2: Ignore local stakeholders at your peril
After working out what it wants to become, and before it commits to a jurisdiction, any startup society needs to understand who’s who in the zoo, so to speak. If the first iron rule is about time, then the second is about power: who holds it, and who can either make or break you. And this is where stakeholder mapping becomes essential. Stakeholder mapping needs to be conducted at all levels, from the local to the national and even the supra-national. Ask: Who has the power to stop our operations, and have we met them yet?
From what I can tell, Malaysia seems a clear illustration of this rule in action. Balaji’s public appeal to Prime Minister Anwar, with a threat to freeze investment attached, was received by Malaysians as a demand that the federal government intervene in what was a state administrative matter. Johor is a sultanate that guards its prerogatives, and its relationship with Putrajaya is sensitive. A stakeholder map would have shown on one page that the entity holding the “off” switch was probably at the state level, and that friendly cabinet-level ministers and agencies at the national level weren’t going to be much help given local political realities, especially given this year’s contentious state elections.
Iron Rule 3: Compliance is not legitimacy
The second iron rule gives you the roster of people who can close you down. The third decides whether anyone in the community will champion your cause and work to keep you around. In jurisdictions with regulatory and governance opacity, you can comply with every law and still run into trouble if you haven't built local legitimacy.
A prospective investor in a network-state-type project once asked me what would prevent a new community from simply buying a chunk of land in the middle of nowhere, where it could just mind its own business and be left alone. I gave two reasons why the invisible approach always backfires.
First, there is no piece of land on the entire big blue Earth that isn't claimed by somebody. The concept of "blank slate" territory is appealing in theory and nonexistent in practice. Inside Network School, the working assumption appeared to be that Forest City is an artificial island*, largely empty, so no Malaysian could feel any attachment to it.
*I suspect it was this type of thinking that spawned Network School’s “island near Singapore” marketing, which seems to have really struck a sour chord with Malaysians on X:
What we found in greater Johor was the reverse. Although nobody in late 2024 knew what Network School was, they definitely had opinions about Forest City. It functioned locally as a monument to how the original Chinese development deal went against Malaysia in general and, if you’ll forgive the term, “majorly screwed us,” in the words of one Johor area business owner. Other Malaysians we spoke with described the island as cursed, and anyone familiar with Malay culture can tell you that word isn't used lightly. Man-made or not, the island represented a huge outlay of both private and public money, and it seemed a blow to their Malaysian national pride that the government would even contemplate selling it off cheaply to foreign investors.
Second, I repeat that, as foreigners, you need a strong reputation-management game from year zero, and invisibility entirely prevents this. Again, with Network School, located on the “island near Singapore,” almost nobody in Johor knew what it was. The vast majority of local Johor residents and Malaysians in general first heard about it in the context of an activist’s highly inflammatory social media post.
A smear campaign is the absolutely worst way for anyone to learn about your project, and this is the whole trouble with invisibility: when the accusation broke, not one person in Johor had a reason to speak up for Network School, because Network School had given too few people in the local area reasons to care whether it stayed or left.
Legitimacy is built from the ground up. It accrues to projects that both ordinary local people and local-level community leadership (e.g., business community, civil society, and the like) have a stake in defending; these types of local defenders are often worth far more than any friendly signature you collect at the national level.
What’s Next for Network School in Kazakhstan?
First, credit where it’s due. The early signals out of Kazakhstan suggest that at least a few of these lessons may be sinking in. In late July, Balaji met Prime Minister Olzhas Bektenov directly, and the plan they described seems like a promising start. The proposed “Network City” near Burabay is explicitly pitched as a partnership with local authorities and the local community; it is meant to be built using modular infrastructure designed to fit the existing landscape (important for preserving the region's pristine nature). Balaji also went out of his way to stress that the whole thing runs on private investment, with no public funds, potentially heading off one of the most dependable sources of local resentment before it even starts. That is what applying the three iron rules from Day 1 looks like.
However, a warm reception from the Prime Minister is precisely the kind of top-level embrace Malaysia extended to Network School, too. So, what should the project keep an eye on? Three risks stand out, each with a recent local precedent.
Land rights. Foreign control of land is one of the most flammable issues in Kazakh politics. In 2016, proposed amendments letting foreigners lease farmland set off nationwide protests (driven largely by fear of Chinese acquisition) and forced the government to shelve the reform entirely. Any campus that takes up real ground is likely to touch on this nerve. Additionally, the Burabay area is one of Kazakhstan’s most beloved natural retreats, so a new “city” going up in it (however modular) invites exactly the kind of scrutiny now trained on foreign developments in protected landscapes elsewhere (also see Albania’s recent “Flamingo Revolution” situation).
Regulatory reversal. Kazakhstan has run the welcome-then-eject cycle for a tech project at least once already. It rolled out the red carpet for bitcoin miners in 2021, and when they strained the grid, the government throttled their power, taxed them, and pushed them out within a year. The lesson here is that a welcome extended by decree can be withdrawn by decree.
Political instability. In January 2022, protests over fuel prices quickly escalated into riots amid what the government called an attempted coup, which required a military intervention of Russian-led CSTO troops, and ended with over 200 dead after Kazakh security forces were given “shoot to kill” orders. Local grievances in Kazakhstan can escalate from zero to a nationwide crisis very quickly, and such political instability puts a visible foreign enclave at elevated risk.
Conclusion
Nobody should take Network School’s exit from Malaysia as a reason to stop building parallel societies, stop iterating, or stop attempting strange and ambitious things in unfamiliar places. In fact, the parallel society movement greatly benefits from more Network School-type experiments, provided they are conducted ethically, which, in this context, means remembering that the host community consists of stakeholders who must not be ignored. As explained at length above, this is for both moral and pragmatic reasons.
Although its first attempt didn’t work out, Network School’s experience in Malaysia should be viewed as the invaluable learning opportunity it is. This is exactly why Polis Labs exists: to research and experiment, knowing that every real experiment carries the risk of failure. After all, the only way to truly waste a failure is to refuse to study it.
So many things to think about as I sit with my tendency to run solo and slide under the radar!
I am also realizing a human is different than their role. Much like a child worth loving is different than their behavior.
When it comes to engaging with a larger whole, it seems the individual and the collective need to learn to dance together without violating either!
Perfectly explain. Hopefully Balaji has learn a lesson and starts being less tone deaf about the social aspect of creating new countries