Associate Professor David Harrich says they have discovered how to modify a protein in HIV so that, instead of replicating, it protects against the infection.
"I consider that this is fighting fire with fire," he said.
"What we've actually done is taken a normal virus protein that the virus needs to grow, and we've changed this protein, so that instead of assisting the virus, it actually impedes virus replication and does it quite strongly."
Audio: Listen to Stephanie Smail's report (AM)
Associate Professor Harrich says the modified protein cannot cure HIV but it has protected human cells from AIDS in the laboratory.
"This therapy is potentially a cure for AIDS," he said.
"So it's not a cure for HIV infection, but it potentially could end the disease.
"So this protein present in immune cells would help to maintain a healthy immune system so patients can handle normal infections."
More than 30,000 people have been diagnosed with HIV in Australia.
If clinical trials are successful, one treatment could be effective enough to replace the multiple therapies they currently need.
"Drug therapy targets individual enzymes or proteins and they have one drug, one protein," Associate Professor Harrich said.
"They have to take two or three drugs, so this would be a single agent that essentially has the same effect.
"So in that respect, this is a world-first agent that's able to stop HIV with a single agent at multiple steps of the virus lifecycle."
He says the new treatment has the potential to make big improvements in the quality of life for those carrying HIV.
"I think what people are looking for is basically a means to go on and live happy and productive lives with as little intrusion as possible," he said.
"You either have to eliminate the virus infection or alternatively you have to eliminate the disease process and that's what this could do, potentially for a very long time."
Professor Harrich says animal trials are due to start this year and early indications are positive.
"This particular study is going to have some hurdles to jump through, but so far every test that we have put this protein through has passed with flying colors," he said.
"This particular year we're moving this into animal models, and based on the preliminary data we have done we expect that this will proceed really quickly."
The research is published in the journal Human Gene Therapy.
