In January 1988, one of Taiwan’s most senior nuclear engineers defected to the United States after passing crucial intelligence on a top-secret program that would alter the course of Taiwan’s history.

Colonel Chang Hsien-yi was a leading figure in Taiwan’s nuclear weapons project, a closely guarded secret between the 1960s and ‘80s, as Taipei raced to develop its first nuclear bomb to keep pace with China.

He was also a CIA informant.

Chang exposed Taiwan’s secret nuclear program to the United States, its closest ally, passing intelligence that ultimately led the US to pressure Taiwan into shutting down the program – which proliferation experts say was near completion.

“I decided to provide information to the CIA because I think it was good for the people of Taiwan,” said the 81-year-old. “Yes, there was political struggle between China and Taiwan, but developing any kind of deadly weapon was nonsense to me.”

Chang’s story bears similarities with that of Mordechai Vanunu, the Israeli whistleblower who famously exposed his country’s clandestine nuclear program to the world. But while Vanunu went public with his country’s progress, Chang’s whistle-blowing was done in secret and without any fanfare.

Taiwan’s nuclear ambitions

In 1964, just 15 years after the Chinese civil war ended with communist victory, leaving Chiang Kai-shek’s nationalists controlling only Taiwan, Beijing successfully tested a nuclear weapon – deeply unsettling the government in Taipei which feared it could one day be used against the island.

Two years later, Chiang launched a clandestine project to lay the technical groundwork for nuclear weapons development over the next seven years. The Chungshan Science Research Institute ran the project under the Defense Ministry, and it was there that Chang began working as an army captain a year later.

He was picked for advanced nuclear training, which would involve stints in the US. After studying physics and nuclear science in Taiwan, he attended Oak Ridge National Laboratory in Tennessee.

Despite Taipei’s official statements that its nuclear research was only for peaceful purposes, Chang said students sent to the US were all aware of their true mission: learning skills for weapons development.

“We know precisely – even though it’s not in the written statement – we know what we are going to do, what kind of area we should concentrate on,” Chang said.

“We were kind of excited and trying to get the job done,” he added. “All we did was focusing on the area they assigned us, we put all our efforts to do it, to learn as much as possible.”

While he was at Oak Ridge, Chang recalled, the CIA already had an interest in him.

“In 1969 or 1970, I remembered receiving a phone call,” he said. The caller said he was “with a company and they are interested in the nuclear power business… they offered to take me out for lunch.”

“At that time, I said I had no interest because I had a mission-oriented assignment. But I was not aware he was from the CIA; I only knew that after quite a few years.”

American suspicions

In 1977, a year after attaining a PhD in nuclear engineering from the University of Tennessee at Knoxville, Chang returned to Taiwan. He was promoted to lieutenant colonel and spearheaded the development of computer codes for simulating nuclear explosions at the Institute of Nuclear Energy Research (INER), a national laboratory that covertly advanced weapons development under civilian pretenses.

Taiwanese leaders faced a delicate balancing act: the United States strongly opposed new nuclear weapons programs anywhere in the world, and Taipei could not afford to alienate its most crucial ally. The US has long relied on nuclear deterrence as part of its broader strategy to counter China’s stockpiling of nuclear warheads. But, under a policy of nonproliferation, it opposes any country newly developing nuclear weapons.

Back then, Taiwan was not the wealthy and vibrant democracy it is today. It was a developing economy under the autocratic rule of the Chinese Nationalist Party, or Kuomintang. That regime continued to hold a seat at the United Nations until 1971, and maintained formal diplomatic relations with the United States until 1979.

To minimize the risk of its nuclear ambitions being exposed, the island only sought to secretly establish the capability to produce nukes quickly at any time, but not build a stockpile.

“Taiwan’s cover stories were unbelievably good,” said David Albright, a nuclear proliferation expert and author of “Taiwan’s Former Nuclear Weapons Program: Nuclear Weapons On-Demand.”

“They always emphasized that the research was only for civil purposes… (US) officials didn’t know how to breach this cover story.”

But the risk of a cross-strait nuclear conflagration weighed on Chang. Chinese leader Deng Xiaoping, who assumed power in 1978, warned that if Taiwan acquired nuclear weapons, China would respond with force.

“I think they’re quite serious,” Chang added. “I believed in that.”

“I didn’t want to have any conflict in any way with mainland China,” he said. “Using any kind of deadly chemical or nuclear weapons… it’s nonsense to me. I believe we are all Chinese and that doesn’t make sense.”

So when CIA agents approached Chang again during a trip to the United States in 1980, he agreed to speak.

“They said, ‘We know you, and we’re interested in you,’ and we had a conversation,” Chang said, adding that the Americans put him through a “very thorough” lie-detector test to ensure he was not a double agent. He assisted the CIA with some ad hoc tasks before becoming an informant in 1984.

For the next four years a CIA case officer, identified only as “Mark,” met with Chang every few months at safehouses around Taipei, including a condo near Shilin Night Market – one of the island’s most famous street food destinations.

In those meetings, the CIA asked him to corroborate intelligence, share information about recent projects at INER, and take photos of sensitive documents.

“All those conversations were quite professional. He would take a pencil and notebook to write down my answers,” Chang said. “He kept saying that they will try their best to keep me and my family safe.”

The Chernobyl disaster in 1986, a catastrophic nuclear accident in Ukraine that exposed hundreds of thousands of people to harmful radiation, solidified Chang’s conviction that halting Taiwan’s nuclear weapons program was imperative.

That same year, Vanunu publicly exposed details of Israel’s clandestine nuclear program, handing what he new to the British media and causing an international sensation. He was later kidnapped by Mossad agents, returned to Israel and prosecuted, spending years in prison.

A new chapter of life

Chang’s life – and those of his wife and three children – took a dramatic turn in January 1988, when the CIA exfiltrated them to the US.

By then, President Ronald Reagan’s administration had amassed sufficient evidence and seized the opportunity created by the death of President Chiang Ching-kuo – Chiang Kai-shek’s son – to pressure his reformist successor Lee Teng-hui into cooperation.

Albright, the expert and author, said Chang was the most crucial informant in arming Washington to shut down the Taiwanese program.

“The United States had been in a cat-and-mouse game with Taiwan over its nuclear program for years,” he said. “Chang really made sure the US had heavy evidence that Taiwan couldn’t deny… and directly confront the Taiwanese.”

In the months after Chang’s departure, the US sent specialists to dismantle a plutonium separation plant – a facility designed to extract nuclear materials for weapons production. The team also oversaw the removal of heavy water, a substance used as a coolant in nuclear reactors, and irradiated fuel, nuclear fuel that can be reprocessed to extract materials for nuclear weapons.

Hero or traitor?

To date, Chang’s decision to work with the CIA has remained controversial in Taiwan, which in the intervening years has continued its massive industrial and economic expansion, becoming a full democracy in the 1990s.

But cross-strait hostilities persist. Taipei has come under growing military pressure from China, which now has the world’s largest military and is becoming more assertive in its territorial claims over Taiwan. The Chinese communist Party has vowed to take Taiwan by force if needed, despite having never controlled it.

Beijing dwarves Taiwan’s military, spending about 13 times more on defense. Some have argued that if Taiwan had successfully acquired nuclear weapons it could have served as an ultimate deterrent – paralleling Ukraine, where Russia might not have invaded if Kyiv had retained its Soviet era nuclear arsenal instead of giving it up.

Some Taiwanese have criticized Chang, saying he overstepped by deciding unilaterally that the island is better off without a nuclear deterrent.

“I believe he is a traitor,” said Alexander Huang, an associate professor in strategic studies at Tamkang University, because the weapons “would be seen as a useful tool in bargaining for a better diplomatic result” with Beijing.

But Su Tzu-yun, a director of Taiwan’s Institute for National Defense and Security Research, said the lack of a nuclear option has not overly affected Taiwan’s modern defense capabilities, because precision ammunition can be used to achieve similar objectives to those of tactical nuclear weapons.

“The Taiwanese government back then thought that if China landed in Taiwan, it could use tactical nuclear weapons to eliminate the landing troops,” he said. “But in their absence, we can also employ precision weapons like missiles to replace them.”

Taiwan buys these weapons from the US, which – despite shutting down the nuclear program – remains its key military partner, supplying ammunition, training, and defense systems.

Besides weaponry, the island has what some consider a more effective deterrent than nuclear bombs. In 1987 – just one year before the nuclear program was shut down – tech entrepreneur Morris Chang founded the Taiwan Semiconductor Manufacturing Company (TSMC), which now produces an estimated 90% of the world’s super-advanced semiconductor chips for tech companies, including Apple and Nvidia.

The island’s integral role in the global semiconductor supply chain, some observers say, would be enough to deter China from launching an invasion, forming what is dubbed its “Silicon Shield.”

Albright, who conducted extensive research into the Taiwanese program, also said its success would not have been beneficial to Taiwan.

“I think [it] would have raised the military risk of a Chinese attack,” he said, while Washington could have also responded by “reducing its security commitment or limiting military aid” once Taiwan’s capabilities were known.

As for Chang Hsien-yi, who became a Christian and enjoyed playing golf outside a part-time role at a nuclear safety consultancy firm, the decision he made four decades ago was correct.

“Maybe that’s good for the Taiwanese people. At least [we] didn’t provoke mainland China in a such way to start an aggressive war against Taiwan,” Chang said.

“I did it with my conscience clear, there is no betrayal – at least not to myself.”

A fish under a roof. A stick figure without a head. A series of lines that look like a garden rake.

These symbols are part of an entirely undeciphered script from a sophisticated ancient civilization thousands of years old. And they remain an enduring mystery that has sparked heated debates, death threats to researchers, and cash prizes for the coveted answer.

The latest such prize was offered last month by the chief minister of one Indian state: $1 million to anyone who can decode the script of the Indus Valley civilization, which stretched across what is now Pakistan and northern India.

“A really important question about the pre-history of South Asia could potentially be settled if we are able to completely decipher the script,” said Rajesh P. N. Rao, a computer science professor at the University of Washington who has worked on it for more than a decade.

If deciphered, the script could offer a glimpse into a Bronze Age civilization believed to rival ancient Egypt and Mesopotamia. Some believe this vast domain held millions of people, with cities that boasted advanced urban planning, standardized weights and measures, and extensive trade routes.

Perhaps more importantly, it might help answer fundamental questions about who the Indus Valley people and their descendants were – a politically fraught debate about the disputed roots of modern India and its indigenous inhabitants.

“Whichever group is trying to claim that civilization would get to claim that they were among the first to have urban planning, this amazing trade, and they were navigating seas to do global trade,” Rao said.

“It has a lot of cachet if you can claim that, ‘Those were our people who were doing that.’”

Why is it so hard to decipher?

Though the script has remained unsolved since its earliest samples were published in 1875, we do know a little about Indus Valley culture itself – thanks to archaeological excavations of major cities like Mohenjo-daro, located in what is now Pakistan’s Sindh province, about 510 kilometers (317 miles) northeast of Karachi.

These cities were designed along a grid system like New York City or Barcelona, and were equipped with drainage and water management systems – features which at that point were “unparalleled in history,” one paper said.

Throughout the second and third millennia BC, Indus merchants traded with people across the Persian Gulf and the Middle East, their ships bringing copper ingots, pearls, spices and ivory. They crafted gold and silver jewelry, and built faraway settlements and colonies.

Eventually around 1800 BC – still more than 1,000 years before the birth of ancient Rome – the civilization collapsed and people migrated to smaller villages. Some believe climate change was the driving factor, with evidence of long droughts, shifting temperatures, and unpredictable rainfall that could have damaged agriculture in those final few centuries.

But what we know about the Indus civilization is limited compared with the wealth of information available about its contemporaries, such as ancient Egypt, Mesopotamia, and the Maya. That is largely because of the undeciphered script, which was found on artifacts such as pottery and stone seals.

There are a few reasons it’s been so hard to decode. First, there aren’t that many artifacts to analyze – archaeologists have only found about 4,000 inscriptions, compared with an estimated 5 million words available in ancient Egyptian, which includes hieroglyphics and other variants.

Many of those Indus relics are very small, often stone seals measuring one square inch – meaning the script on them is short, most sequences containing only four or five symbols.

Crucially, there isn’t yet a bilingual artifact containing both the Indus Valley script and its translation into another language, as the Rosetta Stone does for ancient Egyptian and ancient Greek. And we don’t have clues such as names of recognized Indus rulers that could help crack the script – the way the names of Cleopatra and Ptolemy helped decipher ancient Egyptian.

There are some things that experts largely agree on. Most believe the script was written from right to left, and many speculate it was used for both religious and economic purposes, such as marking items for trade. There are even some interpretations of signs that multiple experts agree on – a headless stick figure representing a person, for instance.

However, until a Rosetta Stone equivalent is found, these remain unproven theories. “No unanimity has been reached even on the basic issues,” wrote Indus experts Jagat Pati Joshi and Asko Parpola in a 1987 book that catalogued hundreds of seals and inscriptions.

Even decades later, “not a single sign is deciphered yet,” said Nisha Yadav, a researcher at the Tata Institute of Fundamental Research in Mumbai, who worked with Rao on the project and has studied the script for nearly 20 years.

Controversial theories

For some people, solving the script isn’t just about intellectual curiosity or academic study – it’s a high-stakes existential question.

That’s because they believe it could settle the controversy of who exactly the Indus people were, and which way migration flowed, in or out of India.

There are two main groups vying to claim the Indus civilization. One group argues the script has links to Indo-European languages such as ancient Sanskrit, which spawned many languages now spoken across northern India.

Most scholars believe Aryan migrants from Central Asia brought Indo-European languages to India. But this group argues it was the other way around – that Sanskrit and its relatives originated in the Indus Valley civilization and spread out toward Europe, said Rao.

He described their claim as: “Everything was within India to begin with … Nothing came from outside.”

Then there’s a second group that believes the script is linked to the Dravidian language family now largely spoken across South India – suggesting Dravidian languages were there first, widely spoken across the region before being pushed out by the arrival of Aryans in the north.

M. K. Stalin, the southern Tamil Nadu state leader offering the $1 million prize, is among those who believe the Indus language was a Dravidian ancestor – which Rao described as the more “traditional” theory, though there are respected scholars on both sides.

Then there are some like Indus expert Iravatham Mahadevan, who argued there’s little point in the debate since the distinction between northern Aryans and southern Dravidians isn’t clear anyway.

“There are no Dravidian people or Aryan people – just like both Pakistanis and Indians are racially very similar,” he said in a 1998 interview.

“We are both the product of a very long period of intermarriage, there have been migrations … You cannot now racially segregate any element of the Indian population.”

Still, the question is fraught. In a 2011 TED Talk, Rao said he received hate mail after publishing some of his findings. Other researchers have described receiving death threats – including Steve Farmer, who along with his colleagues stunned the academic world in 2004 by arguing the Indus script doesn’t represent a language at all, but is merely a set of symbols like those we’d see on modern traffic signs.

How they’re trying to crack it

Despite these tensions, the script has long enamored researchers and amateur enthusiasts, with some dedicating their careers to the conundrum.

Some, like Parpola – one of the eminent experts in the field – have tried figuring out the meaning behind certain signs. For instance, he suggests, in many Dravidian languages the words for “fish” and “star” sound the same, and stars were often used to symbolize deities in other ancient scripts – so Indus symbols that look like fish might represent gods.

Other researchers, like Rao and Yadav, are more focused on finding patterns within the script. To do this, they train computer models to analyze a string of signs – then take away certain signs until the computer can accurately guess what the missing symbols are.

This is useful for several reasons: We can better understand patterns in how the script works – like how the letter “Q” is most often followed by “U” in English – and it can help researchers fill in the gaps for artifacts with damaged or missing signs.

Significantly, knowing these common patterns can help identify sequences that don’t follow the rules. Yadav pointed to seals found in West Asia, far from the Indus Valley; while they used the same Indus signs, they followed entirely different patterns, suggesting the script may have evolved to be used across different languages, similar to the Latin alphabet.

Then there are your average Joes, fans of the puzzle who want to try their hand at solving it. With the announcement of the $1 million prize – though no clear information about where people can apply for it – amateurs have flocked to experts to eagerly share their theories.

“I used to get about one or two emails a week. But now, after the prize was sent out, I pretty much get emails every day,” Rao said. They come from all sorts of people around the world, writing in different languages – with even families working on the puzzle together.

After so many years, Rao swings between optimism and resignation. Any further breakthrough would require international multi-disciplinary teamwork, massive funding, and even political negotiations to allow excavations in border areas disputed by India and Pakistan, he said.

But on good days, he’s still hopeful. So is Yadav, who has been fascinated by the Indus Valley civilization since learning about in the fourth grade. Even without the promise of a solution, the beauty of the task draws her back year after year.

“I look forward to working on the problem every day,” she said. “If we decipher the script, it will open a window into the lives and ideology of Indus people. We will get to know a lot of things about our ancestors … what they were thinking, what were they focused on?”

These details are “just hiding from us today,” she added. “That keeps me glued to the problem rather than anything else.”

The extractor’s part-rusting mechanical arm winds out over the frozen ground, over a sprawling lunar landscape of unnatural colors. The mining of titanium has a greater urgency than ever, here in Irshansk.

The electricity that powers the vast machines is only sometimes on for three hours a day. But resources like titanium are potentially key to the moonshot rare earth minerals deal that is suddenly the focus of talking peace in Ukraine. The deal’s signatories, the United States and Ukraine, appear to have opposing interpretations of its terms, which leave many thorny details for a later discussion.

Some current and former US officials have cast doubt on President Donald Trump’s claim that the potential deal he is on the verge of signing with Ukrainian counterpart Volodymyr Zelensky would offer the US easy access to a plethora of rare earth minerals.

Much of what does exist will be difficult to exploit, particularly at a time of war.

And from this beleaguered mine in the northwestern town of Irshansk, it is hard to see how Ukraine could, in this lifetime, get to the half a trillion dollars Trump has suggested they might repay.

“Now we don’t know what and how our work will go on even tomorrow”, said Dmytro Holik, director of mining and concentrating plant at Ukrainian conglomerate Group DF.

“Every day we see how Ukraine’s energy system is being destroyed. Every day, entire regions are cut off in an emergency,” he added, a reference to the waves of drones and missiles Russia pounds Ukrainian homes and energy infrastructure with each night.

The plant’s staff are mostly men, kept away from conscription as titanium is considered a critical industry. Profits are low, prospects dim. “Our enterprise is now very unstable, and this leads to a very high cost of our products,” Holik said.

The proposed minerals bonanza now at the heart of continued US aid to Ukraine in the largest war in Europe since the 1940s, seems to speak to a fantastical future world of prosperity.

Trump on Thursday held out the possibility of American personnel in Ukraine working to extract minerals once a mineral resources deal – and peace – was in effect.

“When you talk about economic development, we’re going to have a lot of people over there,” he said. “So we’ll be working in the country. So I just don’t think you’re going to have a problem.”

Opaque deal

It refers to half the value of “relevant Ukrainian Government-owned natural resource assets,” but says the specific details will be “agreed by both Participants, as may be further described in the Fund Agreement.” The deal goes on to say these won’t include “current sources of revenues… already part of the budget revenues of Ukraine.”

The extent of Ukraine’s mineral wealth is unclear.

Ukrainian officials accept they sometimes rely on Soviet-era geological dating. Yet in a recent presentation by the Ministry of Ecology and Natural Resources, Kyiv claimed 7% of the world’s titanium production, and to have 3% of the lithium reserves – which have not been mined yet. It also said Ukraine was in the top five nations of graphite reserves, and had deposits of the rare earth minerals tantalum, niobium and beryllium.

The numbers resemble those in the Ukraine chapter of the US Geological Survey’s 2020-2021 Minerals Yearbook, written before Russia’s full-scale invasion disrupted production. The USGS said at the time that Ukraine was the fifth-ranked producer of titanium sponge, and the sixth largest producer of graphite.

A USGS mineral commodity summary for 2025 had no figure for graphite reserves, and said, among other observations, that Ukraine was a source of the rare earth metal scandium and had shuttered a 1.7 million-ton a year alumina refinery since 2022.

Natalia Bariatska, a doctor of geology and member of the Australian Institute of Geoscientists, said critical raw materials were all at different stages of research and exploration.

“It is very difficult to talk about the actual value of these deposits,” Bariatska said. “We can speak about the value of the elements in the subsoil, but we have to understand it takes a lot of investment to extract, process and sell them.”

While the framework deal leaves it unclear what assets will be impacted by any future fund, US National Security Adviser Mike Waltz made an explicit reference during a White House briefing on February 20 to one Ukrainian “aluminum foundry.”

“It’s been damaged, it is not at its current capacity, if restored it would account for America’s entire imports of aluminum for an entire year,” Waltz said.

Waltz did not name the foundry, and the White House did not respond to a request for clarification at the time of publication. But the most likely facility he was referring to is the Zaporizhzhia industrial aluminum integrated plant.

Mothballed a decade ago, a video of the plant released by Ukrainian investigators in 2015 shows it in significant disrepair. It is since running on a much-reduced staff, and has been hit by a missile, according to a filing with the State Property Fund of Ukraine.