Is Kanz al-Ummal byAli ibn Abd-al-Malik al-Hindi a authentic hadith collection? How many weak ahadiths does it have?
Anyone who have basic knowledge of hadees science knows that their are many types of collections of ahadees from which sahih is the type in which the collector/author tries his best to make sure that only authentic ahadees come through his book.
The purpose of the book was to arrange al-Suyuti's al-Jaami` al-sageer into fiqhi sections. Writer of kanz ul ummaal also added some other ahadees too in his book though.
And JAMIA is also a type of hadees collection, like sahih is a type of collection, jamia is a type of collection in which author's main intent is only to collect all the ahadees he can find in one place. Not caring much about authenticity or anything.
Also the writer of this book book does not claim to have only authentic/sahih ahadees in this book though, their are many weak narrations in this book too. So as a whole book, we can't call it authentic like we call sahih bukhari or sahih muslim authentic books.
And i know quite a few zaeef ahadees in this book, but i don't know the exact count though. I hope someone will add the count in this answer.
Allah knows best
To date, no one has authenticated Kanz Al-Umal such that someone could give you a number. However, it is well known to be full of weak reports as the author and editor (Suyuti and Hindi) did not intend to collect only Sahih Hadith. There are two ways of approaching the answer and a buffer. The first is the safe count and the second is the more realistic count.
1) Al-Albani counted when separating Jami As-Saghir 8,202 authentic and 6,450 weak reports totaling 14,652 which is 31,972 reports less than Kanz (note: Jami Saghir is included in Kanz).
2) The compiler of Mawsu'ah Al-Albani As-Sahihah (Hamza Ahmad Az-Zayn, the same Muhaddith who completed Musnad Ahmad when Ahmad Shakir died) counted 27,649 authentic reports and the complier of Mawsu'ah Al-Albani Ad-Da'ifah (Ali Hasan Al-Halabi) counted 31,577 reports, totaling 59,226 which is 12,602 reports more than Suyuti/Hindi collected. However, the largest collection of reports to date is the Mawsu'ah At-Tafsir Bil-Ma'thur by At-Tayyar who, along with a team of 41 Phds and Master students collected 85,730 reports, almost double Kanz Al-Umal. Assuming there is no repetition in the Tafsir, there are 26,504 reports more than what Al-Albani commented on and 39,106 more than Kanz.
Given the figures we can speculate that there are at least 8,200 authentic reports in Kanz but closer to 27,600 and there are at least 6,400 weak reports but closer to 31,500 with a buffer of 26,500 to 39,100 wiggle room between them. Its possible to authenticate the Kanz using the Mawsu'ah Al-Albani As-Sahihah and Mawsu'ah Al-Albani Ad-Da'ifah with 12,600 reports to spare.