Google said it blocked more than 100 bad ads per second last year, releasing a blog post on Wednesday to assure marketers of its efforts to protect the ad ecosystem from ad fraud, malware and content scammers.

“Digital advertising plays an important role in making the web what it is today,” wrote Scott Spencer, Google’s director of sustainable ads. “In order for this ad-supported, free web to work, it needs to be a safe and effective place to learn, create and advertise.”

In his blog post, Spencer stressed the importance of user experience when policing Google ads, maintaining a balance between monetization and honesty. News sites and those posing as news sites were the biggest offenders last year.

Google has cracked down on those who use AdWords to promote clickbait, suspending more than 7,000 accounts last year. The company also blocked over 12,000 websites for “scraping,” the act of duplicating and copying content from other sites.

In 2017, Google removed 320,000 publishers from its ad network for violating publisher policies, the company said, adding that it blacklisted nearly 90,000 websites and 700,000 mobile apps. Ads were removed from 8,700 web pages that violated Google’s terms regarding discrimination and intolerance, as well.

Google is also working to combat trends like gambling, cryptocurrency and scam artists posing as rehabilitation centers online.

“We’re constantly updating our policies as we see new threats emerge,” wrote Spencer.

Earning $27.7 billion in ad revenue for the fourth quarter of 2017 alone, Google has a lot to lose over brand safety concerns and fake news. Along with Twitter and Facebook, Google testified before US Congress in November about possible abuses by Russian agents on their platforms to interfere with State elections.

Google denied any knowledge of such abuses on its platform, but the advertising giant has other, more local problems to deal with. During the company’s fourth-quarter earnings call, chief executive Sundar Pichai assured investors they were working to stop the monetization of offending content on YouTube to include extremist groups and hate speech.