A trove of declassified Office of Naval Intelligence files reviewed at the U.S. National Archives in 2025 upends a durable assumption about Dai Li (戴笠), wartime director of Nationalist China's secret service: that American officers' attitudes toward him mirrored their attitudes toward Chiang Kai-shek (蔣中正). The documents reveal something more interesting — and more instructive — than simple political alignment.
Dai Li headed the Bureau of Investigation and Statistics, known by its Chinese abbreviation Juntong, and served as director of the Sino-American Cooperative Organization (SACO), a joint intelligence program operating under U.S. Navy command. His American counterpart and close personal friend was Rear Admiral Milton E. Miles, SACO's deputy director. Together they ran one of the most operationally effective — and politically contentious — intelligence partnerships of the Pacific War.
The common assumption, widespread among Juntong veterans interviewed in Taiwan, holds that Americans who admired Chiang Kai-shek would naturally have embraced Dai Li, while those who resented Chiang would have opposed him. The archival record shows this framing misses the point entirely.
OSS Opposition Was a Turf War, Not an Ideological Verdict
Among the Naval Intelligence files, a document titled "Responsible American and Allied Officers Who Know General Tai Li" catalogued how various American and Allied commanders regarded Dai Li. Its opening finding is unambiguous: every member of the Office of Strategic Services (OSS) — the wartime precursor to the CIA — either disliked or distrusted him, including OSS Director William J. Donovan himself.
The source of that antagonism was institutional, not ideological. President Franklin D. Roosevelt had designated the OSS to consolidate all American foreign intelligence operations. But Dai Li, with the Navy's backing, insisted that every American intelligence activity on Chinese soil had to flow through SACO's framework — effectively shutting OSS out. When Donovan visited Chongqing in late 1943, the confrontation turned personal. The two sides reportedly threatened to have each other's agents killed.
Yet the OSS officers branded as Dai Li's enemies were no friends of the Chinese Communist Party. Col. Paul Helliwell, who oversaw OSS secret intelligence operations, later helped the CIA acquire Civil Air Transport — Claire Chennault (陳納德)'s airline — and supported Nationalist guerrillas isolated in Burma and Yunnan after the mainland fell. In Beijing's vocabulary, he was a die-hard anti-communist. The SACO files' blanket assertion that all OSS personnel distrusted Dai Li may say more about the U.S. Navy's own service parochialism than about any objective assessment of the man.
Even Gen. Joseph W. Stilwell, whose acrimonious relationship with Chiang Kai-shek is well-documented, maintained workable ties with Miles — because he found SACO useful as a check on OSS's ambitions. The point holds: institutional interest, not political sympathy, was the operative variable.
Chennault Relied on Dai Li's Network and Buried the Debt
Within what was then the U.S. Army Air Forces, attitudes toward Dai Li were equally tangled. Gen. Albert C. Wedemeyer (魏德邁), who replaced Stilwell as the senior American commander in China and harbored a genuine personal antipathy toward the Chinese Communists, kept Dai Li at arm's length — not out of pro-communist sympathy, but out of Army institutional suspicion of the Navy's outsized role in Chinese intelligence. He tried to bring SACO under OSS oversight rather than dissolving the partnership, reflecting the perpetual tension between strategic alignment and service-branch identity.
Gen. Claire Chennault, commander of the "Flying Tigers" and the 14th Air Force, fell on the opposite side of that ledger. SACO provided critical weather and targeting intelligence that measurably improved his bombing operations against Japanese shipping and supply lines. Miles stood up a dedicated 14th Naval Air Group that embedded agents on B-24 bombers from the 308th Bombardment Group to sharpen strike precision. SACO personnel traveling with Juntong guerrilla units also radioed coordinates to 5th Group P-40 fighters conducting strikes against Japanese positions in Hunan. Many American aviators shot down over occupied territory were recovered by Juntong-affiliated forces — a debt Chennault never forgot.
In his posthumous memoirs, however, Chennault deliberately distanced himself from Dai Li, crediting Miles and SACO for operations that Dai Li's network had enabled. He understood that the "secret police chief" label was deeply controversial in the United States, and he chose to insulate his legacy from that association. Supporting Chiang Kai-shek, opposing the Communists, and personally valuing Dai Li were, it turned out, three entirely separable positions — even those who relied on Dai Li operationally did not necessarily want their names linked to his.
Where every other American service hedged, the U.S. Navy committed without reservation. Two successive commanders of the 7th Fleet — Admirals Thomas C. Kinkaid and Charles M. Cooke Jr.— both trusted and supported Dai Li unequivocally. Navy Reserve Lt. Charles G. Dobbins (杜賓斯), writing in "China's Tai Li and the SACO Yanks," credited SACO's weather and coastal surveillance with delivering substantial operational value to Task Force 58 during the Iwo Jima and Okinawa campaigns. Though the Republic of China Navy lacked the capacity to contribute directly to the Pacific's final offensives, SACO ensured China was never entirely absent from the Navy's most consequential battles.
The Navy's own files offer a candid diagnosis of why Dai Li's reputation had soured so badly among OSS officers, Army generals, and State Department officials: he could not speak English. Unlike Soong Mei-ling (宋美齡), who could charm Western officials in fluent American idiom, Dai Li was almost entirely inaccessible without interpreters. That linguistic barrier, more than any substantive policy disagreement, had allowed a distorted image of him to calcify across Washington's bureaucracies.
Acting on this assessment, Navy planners recommended that Dai Li be positioned to take command of China's postwar Navy. Their reasoning was strategic: as a naval chief, Dai Li would be compelled to deepen engagement with the United States; the Navy could use that relationship to gradually reshape Nationalist intelligence culture along more democratic and American-friendly lines. By the war's final phase, the Navy had made a calculated institutional bet that Dai Li was its best lever inside the Republic of China's future security establishment.
Chiang Ching-kuo Would Have Eliminated Dai Li on Taiwan
Dai Li died when his aircraft crashed on March 17, 1946. Miles and other Navy advocates regarded the loss as a strategic setback: only Dai Li had the personal networks and institutional authority to wrest control of the Chinese Navy from the entrenched "Fujian clique" of admirals loyal to Chen Shao-kuan (陳紹寬). Gui Yong-qing (桂永清), a fellow Whampoa Military Academy alumnus who eventually took the naval command, never fulfilled the role the Navy had envisioned for Dai Li — and later died under deeply ambiguous circumstances.
Some Juntong veterans and former U.S. State Department China adviser Miles Yu (余茂春) have argued Dai Li's death was no accident — that OSS and the Chinese Communists collaborated to bring down his plane, clearing the path for Communist victory in the civil war. Hsu finds the claim unpersuasive. The Nationalists lost the civil war for political and social reasons — above all, their failure to address rural land grievances — that no intelligence apparatus, however powerful, could have reversed.
The more telling counterfactual concerns what would have happened to Dai Li had he made it to Taiwan. The evidence, Hsu argues, is not reassuring. Mao Ren-feng (毛人鳳) and Mao Sen (毛森), who ran Juntong's successor organization after Dai Li's death, were both eventually purged by Chiang Ching-kuo (蔣經國) — Chiang Kai-shek's son and political heir — as he consolidated control over Taiwan's security services. Mao Sen, who was Chiang Ching-kuo's own cousin, ended up a fugitive in the United States.
Chiang Ching-kuo, who had studied in Moscow under Stalin's system and returned to China with an abiding instinct for political control, regarded autonomous power centers as existential threats. Dai Li — backed by the full institutional weight of the U.S. Navy, commanding vast personal loyalty networks, and sitting on a spy apparatus that predated Chiang Ching-kuo's political rise — would have been precisely the kind of rival the younger Chiang most feared. The Navy's sponsorship would not have protected him. If anything, it would have deepened Chiang Ching-kuo's suspicion about why the Americans were investing in Dai Li rather than in himself.
For a figure who accumulated that much power in that particular era, the outcomes were rarely natural. The skies over Nanjing on March 17, 1946 may have delivered a merciful verdict.
*The author is a military historian.
Original Article in Chinese
