In early March, just as Covid-19 cases began to uptick in some cities across the U.S., a Boston hospital launched a hotline to triage those needing help. Within days, the hotline became overwhelmed, creating wait times of 30 minutes or more, leading to hang ups.

It wasn’t working.

In response, the team at Partners HealthCare met with Microsoft to build an artificial intelligence (AI) tool that could help triage online. Within days of launch, the tool culled 40,000 patients, providing a first layer of assessment in determining if someone may have the disease or something less threatening.

As the world reacted to Covid, hospitals and healthcare organizations were overrun with potential cases. Like Partners HealthCare, many turned to AI to help. Once a tool that hospitals hesitated to test, some thrust the technology into service. As companies prepare for an easing of the current wave of Covid cases, board and management must weigh whether further investments in the technology could prove useful — especially if another surge of illnesses hit. Even with its successes, though, AI remains limited in its capabilities.

While Covid has highlighted AI’s potential in healthcare, it isn’t to the point where it’s used within clinical settings, Eric Topol, the founder and director of the nonprofit Scripps Research Translational Institute, tells Fortune.

But the ability to assess patients faster can still help companies save on costs, coming at a time when revenues at hospitals have been hampered due to the loss of elective procedures. “The economic impact is profound,” said Nassib Chamoun, CEO of Health Data Analytics Institute. “These businesses have a 2% to 3% margin. They’ve lost a lot of revenue. This is an existential threat to healthcare organizations around the country.”

Many hospitals hadn’t implemented AI prior to the crisis because of data constraints. Privacy concerns meant that finding enough data to feed the algorithms, which produces the results, proved difficult. Even if they were approved, doctors hesitated to use the technology.

But it could take 12 hours to analyze patients results looking at chest X-rays or genetic tests — making triage impossible. In comparison, an AI tool by teleradiology firm Vizyon analyzed patients within 10 minutes at a French hospital. Such effectiveness led to other hospitals implementing the tool, as a first-line defense.

Those seeking better treatment plans are also using AI. Auransa, a San Francisco-based company, is using its technology to map the Covid gene in an effort to develop a smaller list of drugs that could provide a more-optimal treatment strategy. With more wins under AI’s belt, Covid could prove as the moment the technology went mainstream.