The advent of generative AI, large language models and consumer-facing clients or chatbots like ChatGPT are already impacting and inducing shifts in consumer behavior – a factor TV entertainment isn’t immune from, and which could mean changes for how CTV experiences evolve down the line.
Some TV platforms have already started to introduce Large Language Model-powered and AI-based features and efforts to enable more conversational two-way interactions for content search and discovery, such as the addition of Gemini to Google TV devices, and Roku efforts on AI-powered voice search.
And fresh survey data from Nielsen-owned metadata provider Gracenote, released Wednesday, shows how consumers are already and increasingly leaning into AI chatbots, including for entertainment queries, but where concerns over trust and accuracy of results remain.
According to Gracenote’s 2026 GenAI usage study, AI chatbot usage is on the rise, with two-thirds of Americans reporting greater usage today than a year ago. Usage is accelerating particularly among Gen Alpha, the study suggests, with more than half (54%) saying they use AI chatbots every day.
Across age groups, 75% said they use AI chatbots daily or multiple times a week.
While the usage is for a variety of purposes they also pertain to entertainment, where Gracenote found young consumers – particularly Gen Alphas – are tapping AI chatbots as their preferred source for recommendations on TV shows and movies to watch. Nearly half (49%) of Gen Alphas age 13-14 surveyed named AI chatbots as the best source of TV and movie recommendations, outpacing streaming and cable user interfaces and program guides (41%) and internet search engine results (11%).
But it’s not just kids. Gen Z and Millennials, as well as older generations are also using AI to find entertainment and sports information, albeit to a less degree than 13-14-year-olds in the Gen Alpha demo.
Per Gracenote, 33% of Gen Z (age 15-28) use AI to find sports information, increasing to 39% for Millennials (age 29-44) and decreasing to 29% among Gen X (age 45-60). And a good proportion are using AI for content recommendations. About 38% of Gen Z report using AI for that purpose, rising to 41% among Millennials and standing at 32% among Gen X. The entertainment reason each of those three generations tap AI chatbots for most is to find out where a program or a game is playing: Gen Z (39%), Millennials (45%) and Gen X (34%) – speaking to the fragmentation frustration, particularly among sports fans.
The study also asked about consumer preferences for using AI chatbots versus traditional search methods and found that most favor the new technology when it comes to things like complex questions (68% prefer chatbots vs 19% for traditional search), follow-up questions (69% vs. 18%), direct answers (54% vs. 31%) and comprehensive results (50% vs. 30%).
Across the study, chatbots were trusted most for TV and movie recommendations and for helping users find TV, movie and sports programming, at about a quarter of respondents each.
Although more pronounced among Gen Alpha, more than half (52%) of survey respondents said AI chatbot tools could become their preferred way to get information on why, where and when to watch content – with 5% saying they already are.
AI chatbot rise amid content fragmentation, overload
To put some of these findings in the context of entertainment, content discovery on streaming services is a known struggle, with earlier Gracenote data finding it takes people an average of 14 minutes to find something to watch.
And it’s not just a matter of not knowing what to watch. Even when consumers already have a show or movie in mind, with a very fragmented ecosystem of services, platforms and apps, figuring out where that content lives is also a point of friction in the TV viewing, search and discovery experience.
To illustrate the volume, Gracenote’s report showed that as of February 2026, it had a record of more than 1.8 million program titles across nearly 350 SVOD catalogs and almost 210,000 program titles across nearly 2,100 individual FAST channels. These don’t even account for traditional TV channels, which are also part of the content menu across streaming via virtual MVPDs.
And the content deluge isn’t exactly letting up. For example, as the study noted, over the past year, the five major streaming services tracked in the Gracenote Data Hub grew their catalogs a collective 20%.
So although content may very well be king, the amount of both titles and services is becoming a burden for some.
Earlier consumer survey data from Gracenote found 34% of Americans believe the amount of streaming services and content available negatively impacts TV enjoyment, increasing to 48% among those 18-34. And about half of American’s said it’s getting too hard to find content they want to watch because of the sea of services available.
With LLMs and AI chatbots integrated into streaming and TV services, viewers in theory could ask more conversational questions to get to content they want to watch more easily and effectively, and providers stand to benefit from more powerful ranking and sorting capabilities.
And with consumers already turning to AI chatbots for content queries, bringing that utility directly into the living room on the TV screen seems like a somewhat natural next evolution.
AI chatbots in use, but gaps in trust remain
While 66% of overall respondents reported increased use of chatbots for a range of purposes over the past 12-18 months, the study also found a gap between utilizing AI and trusting the results. On that front, traditional search methods still led over AI when it comes to trustworthiness (50% vs 27%) and accuracy (46% vs 33%).
And as Gracenote noted, the survey showed a “somewhat paradoxical” fact in that despite favor for AI chatbots over traditional search for certain purposes, more than 70% of respondents in each age demo (aside from Boomers) fact-check AI chatbot responses, and the vast majority do so by cross-checking with an internet search.
This leads to an emphasis on the need for trusted data underlying AI chat bot responses to queries, particularly those that relate to the specialized world of entertainment data.
Here Gracenote’s findings also help underpin some of the company’s own aims and use of its metadata and recent Video Model Context Protocol (MCP) server product, the latter which can connect a provider’s LLM to the vendor’s data. Gracenote metadata and related tools are helping to ground and validate LLM-based and AI-powered content queries and experiences for TV entertainment providers.
The vendor marked recent deals with Google and Samsung for Gracenote metadata to help power AI use cases on the companies’ respective CTV platforms.
We’ve written a bit about challenges with LLMs in the context of TV entertainment before and using Gracenote’s metadata to help alleviate them, such knowledge-lock, where LLMs doesn’t have access to information outside of its training dates.
LLMs are usually only trained every so often, meaning their data isn’t real-time or continuously updated, so information can be outdated. In addition to LLMs, AI chatbots typically pull from additional data sources to generate strong results, but another issue is hallucinations, where an output might sound correct but isn’t true as AI is non-deterministic in nature and answers aren’t necessarily grounded in current and verified information.
That’s where Gracenote’s structured and refined metadata (as well as products like the MCP server, which doesn’t train LLMs but can connect to them) can come in to serve as a source of truth for TV providers’ underlying models, to essentially check against and validate information or harmonize data – helping to mitigate against knowledge lock and hallucinations and ensure the AI chatbots or clients deliver accurate, up to date and reliable responses to entertainment content queries.
Gracenote held up its recent survey findings as another signal that AI-powered conversational interactions are likely where TV content search and discovery experiences are headed with its metadata, building on comments Gracenote SVP of Product Tyler Bell previously made to StreamTV Insider.
“Our bet is that all players within the CTV space will be using LLMs as the primary mechanism for them and their consumers to interface with media metadata,” Bell told StreamTV Insider in March. “We think this is going to happen very slowly at first and then all at once.”
And the report highlighted some implications of content search and discovery issues for consumers in terms of why streaming services and CTV players should take note.
One, is that younger generations, such as those age 18-34 favor CTV but spend less time with TV – where their average of 16 minutes spent searching could mean less time spent viewing (particularly when many are shifting to prioritize metrics like time spent over the number of users). Second, the risk of churn, where Gracenote reported 54% of 18-34-years as saying they’d consider canceling a service because they couldn’t find something to watch.
"People are rapidly embracing AI as a new way to search, discover and decide what to watch, especially Gen Alpha audiences, who already expect easy-to-use, conversational interfaces,” said Bell in a statement. “But adoption alone is not the story: trust is. The winning platforms will be those that can deliver viewing experiences people can actually rely on — grounded in vetted, timely and high-quality data."