Taboola’s AI answer engine, DeeperDive, is expanding internationally, offering publishers a new way to retain readers on site amid evolving search habits, with notable adoption by USA Today and others across Asia and India.
Publishers are increasingly experimenting with AI tools that keep readers inside their own sites, as referral traffic from search engines becomes harder to rely on. Taboola is trying to ride that shift with DeeperDive, an AI answer engine that sits on publisher pages, fields reader questions and then points them towards relevant archive material and related coverage.
The product has begun spreading beyond its early adopters. According to Taboola, DeeperDive has been taken up by a number of publishers since its late-2025 launch and now supports six languages, with the company saying it reaches nearly 7 million monthly active users. USA Today, one of the first major outlets to deploy it, has seen more than 25 million questions since September, or roughly 1 million a week, while Taboola says publishers using the tool have recorded engagement rates of up to 17%.
USA Today’s product chief, Kara Chiles, said the company is still testing how audiences use the feature and is watching the types of prompts readers choose, especially where they are tied to what they are already reading or to fast-moving news. She said the publisher is treating the tool as a way to learn what audiences want more of, rather than as a simple search box replacement. Taboola founder and chief executive Adam Singolda has said the recirculation rate can exceed 10%, and the company argues that DeeperDive is proving readers will interact with clearly labelled AI features.
The rollout has not been uniform. Taboola said Reach is still in technical integration, despite earlier announcements that it would adopt the product, while HuffPost UK has recently added DeeperDive. Taboola has also been widening the pitch internationally, with press materials citing publishers such as India Today, BuzzFeed Asia and the Bangkok Post, as it tries to position the tool as a global publishing product rather than a narrow experiment for a few US newsrooms.
For Taboola, the AI push sits alongside its long-running recommendation business, which still underpins the company’s publisher relationships. Singolda has said those recommendation units remain central and that the company paid more than $1.5 billion to publishers last year. But with AI answer engines changing how people search and discover information, Taboola is betting that publishers will want a new layer of on-site discovery that can generate both audience data and advertising revenue.
Source Reference Map
Inspired by headline at: [1]
Sources by paragraph:
Source: Noah Wire Services
Noah Fact Check Pro
The draft above was created using the information available at the time the story first
emerged. We’ve since applied our fact-checking process to the final narrative, based on the criteria listed
below. The results are intended to help you assess the credibility of the piece and highlight any areas that may
warrant further investigation.
Freshness check
Score:
8
Notes:
The article was published on May 1, 2026, making it current. However, the content references events from September 2025, indicating that the core information is several months old. This raises concerns about the freshness of the narrative. Additionally, the article appears to be based on a press release from Taboola, which typically warrants a high freshness score but also suggests potential bias and a lack of independent verification.
Quotes check
Score:
6
Notes:
The article includes direct quotes from Kara Chiles, USA Today’s product chief, and Adam Singolda, Taboola’s CEO. However, these quotes cannot be independently verified through other sources, raising concerns about their authenticity and potential bias. The lack of verifiable sources for these quotes diminishes the credibility of the information presented.
Source reliability
Score:
7
Notes:
The article originates from Digiday, a reputable media publication. However, the reliance on Taboola’s press release as a primary source introduces potential bias and raises questions about the independence of the information. The presence of multiple press releases from Taboola further suggests a lack of independent verification.
Plausibility check
Score:
7
Notes:
The claims about the adoption of Taboola’s DeeperDive AI answer engine by publishers like USA Today, HuffPost UK, India Today, BuzzFeed Asia, and the Bangkok Post are plausible and align with known industry trends. However, the lack of independent verification and the reliance on Taboola’s press releases for these claims raise concerns about their accuracy and potential bias.
Overall assessment
Verdict (FAIL, OPEN, PASS): FAIL
Confidence (LOW, MEDIUM, HIGH): MEDIUM
Summary:
The article presents information about Taboola’s DeeperDive AI answer engine, referencing events from September 2025 and relying heavily on Taboola’s press releases. The inclusion of unverifiable quotes and the lack of independent verification sources raise significant concerns about the freshness, originality, and reliability of the content. Given these issues, the article does not meet the necessary standards for publication.