Google is gradually beginning to introduce AI-powered Search Generative Experience (SGE) into the mainstream from the experimental stage.

SGE generates AI-written answers to user queries, especially queries that are too complex to be answered by a single website. It then provides links to the websites referenced in the summary.

The problems started here: in a story first reported by Beeping Computer, SEO consultant Lily Ray noticed that the answers were showing spam and malicious sites.

Being offered in a conversational summary attributed to Google increases the likelihood that someone will be a victim of fraud.

Google says it is constantly updating its algorithm to deal with these types of links, but scammers are also constantly making changes to circumvent protection.

SGE was first launched about a year ago and is available as an opt-in only to users willing to accept the risk and the possibility of being misinformed.

With increasing pressure from new entrants such as Bing and Perplexity, which is fully adopting AI in its search results, Google is offering SGE to more users, including those who have not opted in and may not be aware of the potential for misinformation to be displayed, and They have begun to.

According to Lily Ray, writing in X, the SGE recommends potentially dangerous websites in conversational responses.

If users are unaware of this problem, they are at greater risk of falling for a scam, especially less technically skilled users who may benefit most from AI answers.

Some of these results have been attributed to targeted SEO poisoning campaigns, where scammers use HTML templates that redirect to .online domains and common search terms. This allows them to be indexed by Google even though they contain content that would normally be ignored.

Essentially, the user clicks on a seemingly normal link, such as a sale promotion, and through a series of redirects, eventually reaches a fraudulent site.

Often, they are used to persuade users to accept browser notifications and send them further unwanted advertisements even when they are not visiting the fraudulent site.

According to Bleeping Computer, in rare cases these links can take users to sites that push browser extensions that perform search hijacking with the potential to steal personal information.

The best solution is to remain vigilant of links that seem "too good to be true"; double-check the URL and if it takes you somewhere unprofessional, do not accept the notification and leave the website.

If you click on a malicious link, you should look closely at the domain name of the link to make sure it is legitimate. is the URL spelled correctly? If so, it may be a phishing page and not an official company site.

If you accept the notification and are seeing a lot of advertisements for dubious products and services, go to your browser settings and deactivate browser notifications for that site; in Chrome, this is in Settings, Content and Notifications; in Safari, Settings, Websites, Notifications.

It is also worth investing in the best antivirus program for your computer, especially one that can monitor for malicious code, fake sites, and other content that may spread malware or viruses.

Google says it continues to update its anti-spam system to eliminate these types of search results. The search giant explained that these protections also apply to SGEs and that the examples shared were particularly unusual queries.

This is a particularly significant issue as Google seeks to leverage AI to combat increased competition, since the point of SGE is to handle uncommon and complex queries that often stumble standard search results.