Zentury Spotlight – Google Detects Fake Online Reviews With the Help of AI

Google Detects Fake Online Reviews With the Help of AI

Google is using artificial intelligence (AI) to stop the spread of fake online reviews, which deceive consumers and hurt regional companies.

In 2023, the corporation removed over 170 million bad online reviews, a 45% rise from the year before.

For local business owners who have battled with false online reviews harming their reputations on Google Maps and Search, the crackdown offers much-needed respite.

Google explains in a blog post how its algorithms now swiftly identify questionable review activity by looking for patterns across time. This involves seeing evaluations that are exact replicas of one another on company websites or abrupt jumps in 1- and 5-star ratings.

In the middle of this massive influx of data, Google developed a new algorithm to quickly detect and eliminate false or misleading content regarding online reviews in order to preserve data integrity.

Google reports that a fraud in which people were paid to create fictitious good reviews was discovered by the new algorithm. Google might take measures to shut it down after discovering it. This stopped false reviews from spreading further.

Google also claimed to have blocked or removed the following:

• 170+ million policy-violating reviews,
• 14 million increase in the number of policy-violating videos, 
• 2+ million attempts to make false claims on company profiles have been stopped by improved security measures.

Google claims that by better safeguarding local companies’ online reputations, its ongoing efforts to combat bogus content would protect users and benefit users.

With reputation becoming a more true indicator of service quality and sincere client happiness, the upgrade may result in more even playing fields.

Google Modifies Web Stories Availability

Among other things, Google revised the Web Stories documentation to indicate that Web Stories would no longer be available in Google Images.

Web Stories are a type of brief, visual content meant for quick breaks, such as when taking the bus. They are carousel-style, graphic, and have brief phrases. Swiping opens the next “page” of the content. Google refers to these as visual storytelling and snackable content.

Google Search, which encompasses Google Discover, offers Web Stories. However, they are no longer accessible through Google Images, where they were once found.

The locations of Web Stories have been modified, as indicated by an update to Google’s Search Central developer page.

The new header in the documentation highlights that Web Stories are a component of Google Search. This is the first update to the content.

The content regarding how web stories show up on Google has been removed, and a new header emphasizing that web stories are a search feature has taken its place.

The inclusion of a new paragraph to go with the title above is the biggest change.

Google Explains the Documentation for “Google-Extended” Crawler

Google has clarified the impact on search and included changes in product nomenclature in the documentation of its Google-Extended web crawler user agent. This may raise concerns for individuals who opt to disable the crawler. With regard to limiting content access for AI model training, the revised documentation provides more precise instructions.

Google-Extended was released on September 28, 2023, and provides online publishers with a user agent to manage how their websites are crawled. Publishers have the option to control whether their content is scraped and used in AI training datasets by enabling or disabling the Google-Extended user agent through the Robots Exclusion Protocol.

Google maintains a changelog containing significant revisions to its policies and correspondence with site publishers and the search marketing industry. A modification to the Google-Extended documentation was notified in the developer pages’ changelog on Google.

The update, which follows Bard’s rebranding to Gemini Apps, notes that Vertex AI generative APIs and Gemini Apps now benefit from Google-Extended’s indexing. The revised language minimizes publisher fears about potential repercussions from choosing not to participate in Google-Extended AI data collecting by assuring them that this has no effect on Google Search.

So, what has changed? Google-Extended crawling is limited to Gemini Apps and does not affect Google Search, as made clear by Google’s changelog. The Bard brand name has been replaced with Gemini in the new advice. And the subsequent phrase was added: “Google-Extended does not impact a site’s inclusion or ranking in Google Search.”

Google Confirms Solution for Issue With Search Indexing

The problem with Google Search indexing has been fixed, one week after it was confirmed to be a persistent issue.

Google announced on February 1st that there were delays in their search engine indexing “a small number of sites.” This revelation was made in response to complaints from a number of web administrators and SEO specialists that their most recent content was not being indexed overnight.

According to an update from Google Analyst Gary Illyes on Monday, February 5, “some things take a long time to get fixed properly.” Illyes stated that although the problem had been “mitigated,” there was still a need for a suitable solution.

The Google Search Status Dashboard showed that the indexing problem has been resolved as of February 8 at 3:36 am PST. According to Google, “We resolved the indexing issue.” There won’t be any further releases.

On January 31, at around 3:30 PM PST, indexing issues began. Since then, a large number of news organizations, website publishers, and content producers have complained about difficulties indexing and appearing in Google search results for their recently produced content.

Publishers who were having trouble getting their content indexed and showing up in Google Search may now relax knowing that the issue has been remedied.

Google has not disclosed the reason for the issue. The business stated that in an effort to promptly address the problem, which they accomplished a week later, its engineering teams looked into the underlying reason.