Tech
Man taking picture of Times Square using smart phone, personal perspective view
Getty Images

Ads have entered the chat

Advertisers are crowding into the next digital frontier.

It’s official. Chatbots have entered their ads era.

OpenAI has begun showing ads in ChatGPT and is in talks with The Trade Desk to expand the effort. Google already runs ads across its AI search products, and Adweek recently reported that Google plans to bring them to Gemini this year. (Google has since pushed back on the idea, but it’s extremely hard to imagine Google, which makes the vast majority of its money from ads, abstaining forever.) Most recently, The Information reported that Amazon, which is experimenting with ads inside its AI shopping assistant, Rufus, is also exploring licensing that technology to others.

Startups are all over the space, too. AI ad marketplace Koah, which recently raised $20 million, is placing ads inside everything from coding copilots to AI pediatricians to meal-planning chatbots.

Why?

AI chatbots are the innovation du jour, popping up inside companies and apps of all stripes — and launching as businesses in their own right. They’ve also been around long enough that the bills are coming due. Running large language models is expensive. As chatbots proliferate across the web, their creators are looking for ways to offset those costs — and, in some cases, build entire business models.

While many including ChatGPT are trying subscription models, there’s a ceiling to how many people will opt to pay for those and how much. Advertising, on the other hand, can grow alongside chatbots’ rapidly expanding adoption. It’s the obvious solution.

“Right now everybody’s not caring about revenue too much because they have hundreds of billions of dollars of venture funding to burn,” Koah founder Nic Baird told Sherwood News. “Eventually they have to actually make unit-economics-positive decisions.”

Advertisers, meanwhile, see opportunity. Users don’t just scroll in AI chats; they ask questions and spell out exactly what they’re looking for. That could make conversational AI more valuable than traditional web ads. Users also spend significantly more time on chatbots than in traditional search. For advertisers, that kind of sustained attention is hard to ignore.

The potential market is enormous, and some analysts think it could rival search. Roger Beharry Lall, research director at IDC, estimates that over time, the roughly $300 billion to $400 billion spent annually on search ads could shift toward AI, with one market cannibalizing the other.

“The vast majority of the internet today is ad-sponsored,” he told Sherwood.  “You’re going to see the exact same thing happen in AI.”

That may not be a hard sell to consumers. Despite how expensive LLMs are to run, many people view them the way they view news sites or social media: free. About two-thirds of consumers say advertising is an acceptable way to monetize AI, according to IDC research — including ads embedded in prompt responses.

Still, ads are only just arriving in AI chats. Consumer sentiment could shift once people see what they actually look like.

What ads in chatbots are like now

We’re in the very early innings. As it stands, ads showing up through Koah and in ChatGPT are based on information shared within the chat itself, rather than data scraped from elsewhere on the web.

Both Koah and OpenAI say ads won’t influence the information provided in responses — a separation they argue is essential to maintaining trust. If you ask about running techniques, you might see an ad for sneakers, but the ad won’t cause the chatbot to steer its answer toward that particular brand.

OpenAI executives have previously expressed skepticism about ads in AI assistants, warning that monetization could distort incentives. But the economics appear to be winning, as evidenced by ChatGPT’s entry into the space.

Anthropic, for its part, has so far resisted advertising altogether, positioning itself as more cautious about commercial influence — even as it advertises on the very big ad platform of the Super Bowl.

The ads within each AI chatbot are native to the platform, but labelled as sponsored and visually distinct from the conversation.

“The format design is really a huge part of it. It takes a ton of work to figure out how to make a great sponsored experience,” Baird said. “You could just take a standard display ad and shove it in there, but that hurts the experience and that really hurts user engagement.”

For now, Koah is mostly working with text ads, though Baird says those could evolve into generative video and more dynamic formats.

A mockup of what an ad from Koah’s two-sided ad marketplace would look like within a chatbot.
Koah

IDC’s Beharry Lall expects chatbot ads to start out resembling traditional internet ads, but become more personalized and interactive over time.

“There is a huge opportunity with generative AI and AI technology to be much more refined,” he said.

The frequency with which users see an ad will depend on the chatbot’s owner, but Baird says it could eventually be one ad every three or four queries.

Koah takes a 30% cut of the ad revenue generated through its marketplace. OpenAI, whose ad business Evercore ISI’s Mark Mahaney estimated could generate $25 billion a year by 2030, is charging $60 per thousand views.

Of course, for AI chats to become a major ad channel, the ads have to actually work — and that’s far from proven.

“ The data matters, the time spent matters, the inventory matters, but also the targeting matters, the self-service advertising platform matters, the price matters,” Nate Elliott, an eMarketer analyst who specializes in AI and advertising, said. “A lot of things will have to click for these ads to be useful.”

Will they be more valuable than the search ads they’re replacing?

“They just launched two weeks ago and they’ve got a lot of work to do,” he said. “Anyone who gives you an answer to that is just guessing.”

More Tech

See all Tech
tech
Tom Jones

Prediction markets have, predictably, been given a boost by the summer of sports

Major platforms like Kalshi and Polymarket have seen huge upticks in users of late, thanks in no small part to what’s felt like a recent sporting smorgasbord, with major competitions across hockey, basketball, and soccer soaking up fans’ time (and spending, clearly) at the outset of summer.

While gaming industry groups may not like it, there’s been a huge change in the methods people are using to put money on the big games, with everyone from fortunate NYC bar owners, to a far less fortunate Spanish supporter, turning to prediction markets to try and turn their sports know-how into cold, hard cash.

According to a new report from Adam Blacker for apptopia, that shift might have been even more seismic than imagined in the wake of the NBA and NHL finals and around the 2026 World Cup kicking off.

While gaming industry groups may not like it, there’s been a huge change in the methods people are using to put money on the big games, with everyone from fortunate NYC bar owners, to a far less fortunate Spanish supporter, turning to prediction markets to try and turn their sports know-how into cold, hard cash.

According to a new report from Adam Blacker for apptopia, that shift might have been even more seismic than imagined in the wake of the NBA and NHL finals and around the 2026 World Cup kicking off.

South by Southwest Conference and Festivals

Gold Tesla Cybercabs are piling up, but they’re not picking up passengers yet

Low-volume production started in April. Now people are noticing them more and more in the wild.

Rani Molla6/15/26
tech
Jon Keegan

Anthropic pulls Fable and Mythos access worldwide after Trump administration bars their use by foreign nationals

Only days after releasing two versions of its next-gen AI model, Anthropic has disabled them for users worldwide.

Anthropic says it received a Friday night order from the Trump administration to suspend access to the models for any foreign national (anywhere in the world) — a group that included some Anthropic employees. In response, the company turned off access to everyone.

Last week, the company released to the public its much-anticipated Claude Fable 5 model (and its restricted version Claude Mythos 5, which is still being tested with trusted partners). Anthropic said in a blog post announcing the action that officials cited national security concerns with the new models, while offering few specific details.

The post said that the government gave the company “verbal evidence of a potential narrow, non-universal jailbreak” of the public Fable 5 model. A jailbreak is a means by which users can evade restrictions built into the code to unlock prohibited functionality. Anthropic downplayed the significance of the attack, and said other major models, such as OpenAI’s GPT-5.5, could also be affected by the technique described.

Fears of these first Mythos-class models being misused are running high, after Anthropic warned the cybersecurity world in May that the advanced cyber capabilities of Mythos have rapidly discovered thousands of vulnerabilities in ubiquitous software, leading to the decision to restrict the full version of the model to a close group of trusted partners for testing.

This morning, Axios reported that Anthropic technical staff have flown to Washington to meet with White House officials to resolve the issue.

The Wall Street Journal is reporting that the Trump administration’s decision to take action against Anthropic was prompted by discussions that Amazon CEO Andy Jassy had with officials, including Treasury Secretary Scott Bessent. According to the report, Amazon researchers said they had been able to evade some of Fable 5’s security restrictions using specific prompts. Amazon is a major investor in Anthropic.

Anthropic is currently suing the US government to fight the Pentagon’s blacklisting of the company on national security grounds.

Last week, the company released to the public its much-anticipated Claude Fable 5 model (and its restricted version Claude Mythos 5, which is still being tested with trusted partners). Anthropic said in a blog post announcing the action that officials cited national security concerns with the new models, while offering few specific details.

The post said that the government gave the company “verbal evidence of a potential narrow, non-universal jailbreak” of the public Fable 5 model. A jailbreak is a means by which users can evade restrictions built into the code to unlock prohibited functionality. Anthropic downplayed the significance of the attack, and said other major models, such as OpenAI’s GPT-5.5, could also be affected by the technique described.

Fears of these first Mythos-class models being misused are running high, after Anthropic warned the cybersecurity world in May that the advanced cyber capabilities of Mythos have rapidly discovered thousands of vulnerabilities in ubiquitous software, leading to the decision to restrict the full version of the model to a close group of trusted partners for testing.

This morning, Axios reported that Anthropic technical staff have flown to Washington to meet with White House officials to resolve the issue.

The Wall Street Journal is reporting that the Trump administration’s decision to take action against Anthropic was prompted by discussions that Amazon CEO Andy Jassy had with officials, including Treasury Secretary Scott Bessent. According to the report, Amazon researchers said they had been able to evade some of Fable 5’s security restrictions using specific prompts. Amazon is a major investor in Anthropic.

Anthropic is currently suing the US government to fight the Pentagon’s blacklisting of the company on national security grounds.

Latest Stories

Sherwood Media, LLC and Chartr Limited produce fresh and unique perspectives on topical financial news and are fully owned subsidiaries of Robinhood Markets, Inc., and any views expressed here do not necessarily reflect the views of any other Robinhood affiliate, including Robinhood Markets, Inc., Robinhood Financial LLC, Robinhood Securities, LLC, Robinhood Crypto, LLC, Robinhood Money, LLC, Robinhood U.K. Ltd, Robinhood Derivatives, LLC, Robinhood Gold, LLC, Robinhood Asset Management, LLC, Robinhood Credit, Inc., Robinhood Ventures DE, LLC and, where applicable, its managed investment vehicles.