Colonel Masanobu Tsuji was a fanatical Japanese militarist and brutal warrior, hunted after World War II for massacres of Chinese civilians and complicity in the Bataan Death March.
And then he became a US spy.
Newly declassified CIA records, released by the US National Archives document more fully than ever how Tsuji and other suspected Japanese war criminals were recruited by US intelligence in the early days of the Cold War.
The documents also show how ineffective the effort was, in the CIA's view.
The records, declassified in 2005 and 2006 under an act of Congress in tandem with Nazi war crime-related files, fill in many of the blanks in the previously spotty documentation of the occupation authority's intelligence arm and its involvement with Japanese ultra-nationalists and war criminals, historians say.
In addition to Tsuji, who escaped Allied prosecution and was elected to parliament in the 1950s, conspicuous figures in US-funded operations included mob boss and war profiteer Yoshio Kodama, and Takushiro Hattori, former private secretary to Hideki Tojo, the wartime prime minister hanged as a war criminal in 1948.
The CIA also cast a harsh eye on its counterparts — and institutional rivals — at G-2, the occupation's intelligence arm, providing evidence for the first time that the Japanese operatives often bilked gullible American patrons, passing on useless intelligence and using their US ties to boost smuggling operations and further their efforts to resurrect a militarist Japan.
The assessments in the files are far from uniform. They show evidence that other US agencies, such as the Air Force, were also looking into using some of the same people as spies, and that the CIA itself had contacts with former Japanese war criminals. Some CIA reports gave passing grades to the G-2 contacts' intelligence potential.
But on balance, the reports were negative, and historians say there is scant documentary evidence from occupation authorities to contradict the CIA assessment.
The files, hundreds of pages of which were obtained last month by the Associated Press, depict operations that were deeply flawed by agents' lack of expertise, rivalries and shifting alliances between competing groups, and Japanese operatives' overriding interest in right-wing activities and money rather than US security aims.
“Frequently they resorted to padding or outright fabrication of information for the purposes of prestige or profit,” a 1951 CIA assessment said of the agents. “The postwar era in Japan ... produced a phenomenal increase in the number of these worthless information brokers, intelligence informants and agents.”
The contacts in Japan mirror similar efforts in postwar Germany by the Americans to glean intelligence on the Soviet Union from ex-Nazis. But historians say a major contrast is the ineffectiveness of the Japanese operations.
The main aims were to spy on Communists inside Japan, place agents in Soviet and North Korean territory, and use Japanese mercenaries to bolster Taiwanese defenses against the triumphant Communist forces in mainland China.
Some of the missions detailed by the CIA papers, however, bordered on the comical.
The Americans, for instance, provided money for a boat to infiltrate Japanese agents into the Soviet island of Sakhalin — but the money, boat and agents apparently disappeared, one report said.
In Taiwan, the Japanese traded recruits for shiploads of bananas to sell on the black market back home.
The operatives also were suspected of having murky links with the Communists they were assigned to undermine, the documents say. The CIA also said some agents sold the same information to different US contacts, increasing their earnings, and funneled information on the American military back into the Japanese nationalist underground.
The files and historians strongly suggest that American lack of knowledge about Japan or interest in war crimes committed in Asia, and a reliance on operatives' own assessment of their intelligence skills, made US officials, in the words of one CIA report, “easy to fool for a time.”
“This was a bunch of Japanese nationalists taking the G-2 for a ride,” said Carol Gluck, a specialist in Japanese history at Columbia University and adviser to the archives working group administering the papers. “One thing that was interesting was how absolutely nonsensical it was, of no use to anybody but the people involved. Almost funny in a way.”
The informants, many of whom were held as war criminals after Tokyo's surrender and subsequently released, operated under the patronage of Major General Charles Willoughby, a German-born, monocle-wearing admirer of Mussolini, a staunch anti-Communist and, as the chief of G-2 in the occupation government, second in power only to his boss, General Douglas MacArthur.
Some of Willoughby's proteges were seen as prime war trial material by Allied prosecutors.
But even as the occupation authorities were recrafting Japan into a democracy, their focus was shifting to containing the Soviets.
Willoughby saw the military men as key to making Japan an anti-Communist bulwark in Asia — and ensuring that Tokyo would rapidly rearm, this time as a US ally.
Historians long ago concluded that the Allies turned a blind eye to many Japanese war crimes, particularly those committed against other Asians, as fighting communism became the West's priority.
Chief among the Japanese operatives was Seizo Arisue, Japan's intelligence chief at the end of the war. Arisue had been a key figure in the pro-war camp and in forging Japan's alliance with Nazi Germany and Fascist Italy in the 1930s.
According to the files, Arisue was soon ensconced in G-2, working with former Lieutenant General Yorashiro Kawabe, who was a military intelligence officer in China in 1938 — to organize groups of veterans and others for underground operations.
These groups consisted of former war buddies and often retained the same chains of command and militarist ideology of the war machine that ground much of Asia into submission in the 1930s and 1940s.
“It shows how we acquiesced to the Japanese ... in order to continue to build up Japan as our ally,” said Linda Goetz Holmes, author of Unjust Enrichment: How Japan's Companies Built Postwar Fortunes Using American POWs. “The whole thing was Cold War fear and an awful lot of postwar compensation issues ... all of that was subservient to our total fear of Russia,” said Holmes, also a historical adviser to the National Archives.
Indeed, that new focus brought some of Japan's most notorious wartime killers under US sponsorship.
Tsuji, for instance, was wanted for involvement in the Bataan Death March of early 1942, in which thousands of Americans and Filipinos perished, and for allegedly co-signing an order to massacre anti-Japanese Chinese merchants in Malaya.
Yet none of that seemed to matter much to American intelligence.
The US Air Force attempted unsuccessfully to recruit him after he was taken off the war crimes list in 1949 and came out of hiding, and CIA and US Army files show him working for G-2. In the 1950s he was elected to Japan's parliament. He vanished in Laos in 1961 and was never seen again.
The Army considered him a potentially valuable source, but the CIA was not impressed with Tsuji's skills as an agent. The files show he was far more concerned with furthering various right-wing causes and basking in publicity generated by controversial political statements.
“In either politics or intelligence work, he is hopelessly lost both by reason of personality and lack of experience,” said a CIA assessment from 1954. Another 1954 file says: “Tsuji is the type of man who, given the chance, would start World War III without any misgivings.”
Kodama was another unsavory player. A virulent anti-communist and superbly connected smuggler and political fixer, Kodama commanded a vast network of black marketeers and former Japanese secret police agents in East Asia.
The CIA, however, concluded he was much more concerned about making money than furthering US interests. A gangland boss, he later played a major role in the Lockheed Scandal, one of the country's biggest post-World War II bribery cases. He died in 1984.
“Kodama Yoshio's value as an intelligence operative is virtually nil,” says a particularly harsh 1953 CIA report. “He is a professional liar, gangster, charlatan and outright thief ... . Kodama is completely incapable of intelligence operations, and has no interest in anything but the profits.”
Nowadays, the most powerful legacy of the US occupation is the democratic freedoms and pacifism built into Japan's 1947 constitution. But the US association with Japanese war criminals illustrates how Washington embraced nationalist and conservative forces after World War II, helping them reassert their grip on the government once the occupation ended in 1952.
“It's hard to imagine back in those days how intent the US was on rapid remilitarization of Japan,” said John Dower, historian and author of Embracing Defeat: Japan in the Wake of World War II.
“When we talk about the emergence of neo-nationalism or a strong right wing in Japan today, this has very deep roots and it involves a very strong element of American support,” he said.
Yet the ex-war criminals failed to rebuild a militarist Japan.
“Prewar right-wing activists who escaped war crime charges in fact did not have much influence in the postwar period,” said Eiji Takemae, historian and author of The Allied Occupation of Japan.
To the Americans, he said, “they were in fact not very useful.”
On the Net:
US National Archives and Records Administration: www.archives.gov/iwg/japanese-war-crimes/introductory-essays.pdf and www.archives.gov/iwg/reports/japanese-interim-report-march-200 2-1.htm
Oct. 27 to Nov. 2 Over a breakfast of soymilk and fried dough costing less than NT$400, seven officials and engineers agreed on a NT$400 million plan — unaware that it would mark the beginning of Taiwan’s semiconductor empire. It was a cold February morning in 1974. Gathered at the unassuming shop were Economics minister Sun Yun-hsuan (孫運璿), director-general of Transportation and Communications Kao Yu-shu (高玉樹), Industrial Technology Research Institute (ITRI) president Wang Chao-chen (王兆振), Telecommunications Laboratories director Kang Pao-huang (康寶煌), Executive Yuan secretary-general Fei Hua (費驊), director-general of Telecommunications Fang Hsien-chi (方賢齊) and Radio Corporation of America (RCA) Laboratories director Pan

President William Lai (賴清德) has championed Taiwan as an “AI Island” — an artificial intelligence (AI) hub powering the global tech economy. But without major shifts in talent, funding and strategic direction, this vision risks becoming a static fortress: indispensable, yet immobile and vulnerable. It’s time to reframe Taiwan’s ambition. Time to move from a resource-rich AI island to an AI Armada. Why change metaphors? Because choosing the right metaphor shapes both understanding and strategy. The “AI Island” frames our national ambition as a static fortress that, while valuable, is still vulnerable and reactive. Shifting our metaphor to an “AI Armada”
The consensus on the Chinese Nationalist Party (KMT) chair race is that Cheng Li-wun (鄭麗文) ran a populist, ideological back-to-basics campaign and soundly defeated former Taipei mayor Hau Lung-bin (郝龍斌), the candidate backed by the big institutional players. Cheng tapped into a wave of popular enthusiasm within the KMT, while the institutional players’ get-out-the-vote abilities fell flat, suggesting their power has weakened significantly. Yet, a closer look at the race paints a more complicated picture, raising questions about some analysts’ conclusions, including my own. TURNOUT Here is a surprising statistic: Turnout was 130,678, or 39.46 percent of the 331,145 eligible party

The older you get, and the more obsessed with your health, the more it feels as if life comes down to numbers: how many more years you can expect; your lean body mass; your percentage of visceral fat; how dense your bones are; how many kilos you can squat; how long you can deadhang; how often you still do it; your levels of LDL and HDL cholesterol; your resting heart rate; your overnight blood oxygen level; how quickly you can run; how many steps you do in a day; how many hours you sleep; how fast you are shrinking; how