The frequency of stories about pedestrians and cyclists being killed by vehicles has become alarming. The sad loss of drivers and passengers used to dominate fatal-accident stories, but now risk seems to be shifting toward those outside vehicles.

At the same time, many years of progress in reducing auto fatalities is being reversed. More people are being killed on U.S. roads. Measures are being taken to combat the rise, but they’re not enough.

Highway fatalities last year increased 7.2 percent over the year before, according to data compiled by the National Highway Traffic Safety Administration, the largest increase in half a century. In the first six months of this year, fatalities are up another 10.4 percent from 2015.

People walking along a shoulder, trying to cross a road or riding a bicycle are disproportionately bearing the brunt of this increased carnage. Their fatalities increased an estimated 10 percent from 2014 to last year, according to the Governors Highway Safety Association.

Pedestrian and cyclist deaths were rising even when overall traffic deaths were still falling. From 2009 to 2014, total traffic fatalities fell 4 percent, but pedestrian and cyclist deaths increased 19 percent.

Car safety improvements have greatly benefited vehicle occupants, helping them survive crashes, but they’ve done relatively little for crash victims outside.

That’s starting to change with increasing use of auto-braking systems in new vehicles. These can sense when the vehicle is about to hit a person or object ahead of it and stop the car without the attention or effort of the driver.

While promising, the overall benefit of this technology is yet to be determined. Even if the systems can recognize nearly all of the pedestrians and cyclists in at-risk situations, the laws of physics and stopping distances would still apply. People will still be hit by auto-braking cars sometimes, and it doesn’t take much for a 2-ton vehicle to kill someone.

There also may be unanticipated effects from auto-braking, such as drivers coming to rely on it and paying less attention to the road. That might lead to other kinds of accidents and fatalities.

And even if auto-braking is a substantial help, it will take decades of people replacing their cars to convert the nation’s vehicle fleet to the technology.

Driver attention already has reached unacceptably low levels. New Jersey State Police say distracted driving has been the No. 1 cause of fatalities for the last six years.

Worse is on the way. In a survey by AAA this year, 42 percent reported reading a text or email while driving. Last year, 30 percent of drivers admitted in a Harris Poll that they text while driving.

In a recent NHTSA survey, a fifth of drivers ages 18 to 20 and almost 30 percent of those 21 to 34 said texting doesn’t affect their driving.

This is similar to the way many people were blind to the risk of driving while intoxicated in the years before widespread public education and aggressive law enforcement. Those efforts have cut drunken driving fatalities in half since 1982.

Campaigns to convince the public of the dangers of distracted driving, particularly involving the use of mobile phones, are becoming common and helping. Ocean City handed out warnings to drivers for three weeks in the spring as part of the state’s U Drive U Text U Pay campaign.

We think reversing the increase of roadway killings will take substantially greater enforcement too. Municipalities using plainclothes spotters, for example, have been far more effective in ticketing drivers on hand-held phones.

Police are also more likely to investigate digital distractions in accident cases. We think that should be part of every crash fatality investigation.

Even all this is unlikely to be enough. Technology help from smartphone makers and wireless carriers, as well as auto insurance revisions, will probably be needed before this battle is won.

A new epidemic of highway deaths seems to be in its early stages, and the sooner it’s treated as such, the easier getting it under control will be.