"The reality is, at the moment, they are not useful," said Professor Carola Vinuesa, one of the authors of the government report and co-director of the NHMRC Centre for Personalised Immunology at the Australian National University.
"At the moment, the quality does not seem to be good enough for these tests to be deployed in large scale."
"The sensitivity is not very good. They are not useful in being able to say 'you were infected'."
Professor Vinuesa said no test currently developed was accurate enough to reliably detect antibodies.
"Most individual results will be false positives," she said. "You cannot have most positive results being false."
A spokesman later said the government ended up buying only 1 million tests, after a third supply contract was cancelled.
"These tests have been purchased as they may have a role to play in population level serosurveillance," the spokesman said.
The current COVID-19 test, known as PCR, can only tell if a person is currently infected. Antibody tests are considered crucial by epidemiologists because they will give, for the first time, a true estimate of the number of people who have caught COVID.
But they are extremely difficult to make 100 per cent accurate.
Antibody levels can vary significantly between people, with some having very low levels. In addition, the tests may also pick up antibodies to other viruses that are similar to COVID-19.
"That gives the potential we are identifying people who have just been exposed to the common cold," said Professor Ivo Mueller, head of immunity at the Walter and Eliza Hall Institute.
Professor Mueller's team is currently trying to make their own accurate antibody test - a process that is proving difficult.