The chatbot wars entered the Super Bowl this year.

At Super Bowl LX, a ChatGPT competitor paid millions of dollars for commercials mocking the leading artificial intelligence chatbot’s plans to put advertisements in its chats.

One of the ads, titled “Betrayal,” showed a man seeking help to communicate better with his mother. His therapist, representing a sponsored bot, offers advice on mending the relationship, then suddenly suggests a mature dating site to connect with “roaring cougars.”

The ads from Anthropic, which has a chatbot named Claude, ends with the tagline: “Ads are coming to AI. But not to Claude.”

AI companies are spending hundreds of billions of dollars and need to generate more revenue to keep spending. Though much of the money comes from subscriptions from companies and other heavy users, companies serving regular consumers will probably need to increasingly rely on ads and other methods to monetize mass market users.

The Super Bowl Sunday ads launched a debate about what a future would look like in which the bots many people talk to all day start pitching products.

OpenAI, which has more than 800 million users, generated around $20 billion in revenue in 2025, according to its chief executive, Sam Altman. That still isn’t enough to cover what it has borrowed and plans to spend.

Last month, OpenAI said it will be testing ads for its free-tier users and its low-cost ChatGPT Go subscribers in the U.S.

“Subscriptions cover the committed users,” said former Google executive Justin Inman, who is the founder of Emberos, a startup that researches brand visibility in AI. “But they have a ton of free users as well.”

Ads have just started rolling out on ChatGPT, and the company has shared examples of what they look like in a chat.

One example showed a static link to purchase hot sauce at the bottom of the answer, labeled ‘sponsored’. Another was more conversational. After answering a user query about Santa Fe, the chatbox provided a link to a desert cottage in the locality.

OpenAI underlined that the ads won’t influence ChatGPT’s answers and will be separate and clearly labeled.

Altman responded to the Anthropic commercial on X, calling it funny but “dishonest.”

“We would obviously never run ads in the way Anthropic depicts them,” he said. “We are not stupid and we know our users would reject that.”

He suggested Anthropic was being elitist.

“Anthropic serves an expensive product to rich people,” he said, while OpenAI feels “strongly that we need to bring AI to billions of people who can’t pay for subscriptions.”

Anthropic was founded in 2021 by former OpenAI employees. Though the two companies have been long-term rivals, the Super Bowl ad was one of the first times the scuffle was so public.

While ChatGPT targeted everyday users, Anthropic has focused on selling chatbot services to business customers. The company has witnessed explosive growth, clocking a reported $9 billion in revenue in 2025, and is projected to reach $26 billion this year.

Demis Hassabis, the CEO of Google DeepMind, which operates Gemini, said in a recent interview that he was surprised by OpenAI’s decision to monetize the chatbot through ads this early. Pushing products mid-conversation inside a chatbot could hurt users’ trust in AI as a helpful assistant, he said.

Though Google’s Gemini chatbot doesn’t push ads, last year the company introduced ads in the AI-generated summaries users see atop Google search results. The company also began testing ads in “AI Mode,” a conversation feature on the Google homepage, where sponsored cards appear below the AI-generated search results.

Elon Musk’s Grok, the AI that is integrated into the platform X, also told advertisers last year that it would start testing ads inside chatbot responses as a way to boost revenue and pay for the expensive chips powering AI.

More U.S. shoppers are already turning to AI chatbots, and a Deloitte survey found that trust in generative AI has been steadily increasing. Younger shoppers are using chatbots for comparison shopping, finding deals, summarizing product reviews, and generating shopping lists.

Even without bribing the bots to provide direct advertising, brands are already trying to find ways to get into the good books of AI search results. An entire cottage industry of startups and consultants has emerged to help retailers and brands ensure their products appear in AI search results, a field called Generative Engine Optimization.

The market for traditional search engine optimization was $20 billion to $25 billion, but the potential for AI-driven commerce is much larger, said Amay Aggarwal, a co-founder of Anglera. His company helped Los Angeles-based e-bike and outdoor goods retailer Retrospec adapt its product catalog so that AI chatbots such as ChatGPT and Gemini could accurately recommend the right bikes for specific conditions.

Even as advertising evolves to embrace AI, many of the top AI companies saw value in old-school Super Bowl television ads. In the era of fragmented internet culture, the Super Bowl remains one of the last major shared American television viewing events that draws more than 100 million viewers. AI companies paid up to $10 million for a 30-second spot.

Super Bowl LX was overrun with advertisements from many AI majors, including OpenAI, which promoted its coding platform Codex, and Google’s Gemini, which spotlighted its photo-generation capabilities.

Despite being the “AI Super Bowl,” none of the major AI companies — OpenAI, Google, Anthropic — made the top 20 brands that performed well in generative AI search and conversation during Super Bowl week.

“Being an AI brand doesn’t automatically translate into being remembered by AI,” said Inman of Emberos, whose company produced The AI Influence Index, which tracked the top seven Super Bowl advertisers and how they were showing up in AI queries.

The seven brands that dominated chatbot searches were XFINITY, Bud Light, Squarespace, Ramp, Budweiser, Volkswagen and Dove.

“As ads move into chatbots, the real competition won’t be for attention — it’ll be for how clearly your message survives retelling by AI,” Inman said.

The people of Orange Cove in Fresno County could soon be an unwilling part of an experiment in dangerous, expensive utility boondoggles. And if California’s gas companies get their way, families statewide will be forced to pay higher energy bills, breathe more indoor air pollution and bear greater safety risks.

Southern California Gas Co. wants to use Orange Cove to test blending hydrogen with natural gas in its pipeline network. This might sound futuristic and clean because it would reduce fossil fuel use, but it would waste $64 million in SoCalGas customer money and threaten this community’s health and safety — without actually fighting climate change.

Worse yet, SoCalGas and two other utilities just petitioned state regulators to skip pilot projects altogether. If approved, they could then request to pump a 5% hydrogen blend across California without demonstrating safety.

The problem is blending hydrogen into pipelines and appliances designed for gas. Hydrogen is leakier and more flammable, and it burns hotter and faster than gas. It can’t be smelled or seen, and burning it increases asthma-causing air pollution in homes and risks damaging appliances. Forcing consumers to burn hydrogen worsens fire, explosion and health risks in our homes, where we should feel most safe

The truth is gas utilities’ hydrogen blending proposals intend to keep customers hooked on pipelines. Utilities earn huge profits on infrastructure investment — over 10% for SoCalGas. The wiser approach for Californians would be to switch from gas to electric appliances, protecting customers from volatile gas prices and toxic indoor air. But that would hurt gas utility profits.

In my state of Colorado, our largest utility, Xcel Energy, proposed mixing hydrogen into the natural gas system serving a Denver suburb. When the community learned Xcel was forcing residents into a dangerous, expensive gas alternative disguised as climate action, they pushed back with enough time to force Xcel to pause its effort.

This story is playing out across the country and the world. In Eugene, Ore., backlash from residents made NW Natural cancel its hydrogen blending pilot. In Massachusetts, state regulators prevented utilities from pursuing similar plans. In the United Kingdom, residents of Whitby and Redcar protected themselves from even larger proposals.

Orange Cove is the next flare-up. SoCalGas began campaigning to blend hydrogen in 2022, but residents recently uncovered the truth and are speaking out accordingly. State regulators are expected to act by June, and their decision will have far-reaching consequences.

SoCalGas’ proposal stems from state policy to slash climate pollution from gas utility systems — a good idea, but a threat to utility profits. In theory, replacing natural gas with hydrogen can help gas utilities cut emissions while still investing in pipelines, because hydrogen can be produced and burned without emitting greenhouse gases.

But that’s where hydrogen’s advantages end.

Let’s air out the proposal’s dirty laundry: SoCalGas’ proposal to blend less than 5% hydrogen into Orange Cove’s system — which serves about 2,000 customer gas meters — would cost $64 million over 18 months. That’s comparable to removing the tailpipe pollution of 100 cars for one year.

That same $64 million could permanently remove the pollution of 12 times as many gasoline cars if used to purchase new electric vehicles. It’s also worth around $32,000 per customer gas meter in Orange Cove — more than enough for the community to install electric heat pumps, heat pump water heaters and induction stoves, zeroing out gas use.

Using that $64 million to fund incentives for cleaner, efficient electric appliances could help tens of thousands of Californians eliminate indoor air pollution and climate emissions.

This price tag is ludicrous for an 18-month experiment. Clean hydrogen is an extremely expensive way to heat homes. Current prices are 10 to 25 times higher than that of natural gas, and even the most optimistic forecasts expect it to remain much more expensive for decades.

Gas utilities claim Orange Cove will “inform the feasibility of developing a hydrogen injection standard” to decarbonize their broader systems, but that hides the truth: Hydrogen blending is a dead end that at best would reduce gas utility climate emissions by less than 7%. California’s gas system was not designed to safely handle more than a small share of hydrogen, so this pilot project couldn’t meaningfully scale up without the wholesale replacement of all gas pipelines and appliances.

Pilot projects seem small in the grand scheme of things, but they lend legitimacy to a bad idea debunked as a climate solution and wisely rejected by other communities time and time again. It would be even worse to ditch pilot tests and skip right to harming Californians with statewide blending.

Hydrogen is not categorically a “false solution” for climate. We need it to clean up things like fertilizer, chemicals and aviation fuel — products without cheaper clean alternatives that are made in specialized industrial complexes overseen by trained technicians.

But California doesn’t need hydrogen to clean up its buildings. Families are already choosing electric appliances for higher-quality, fully clean service. Hydrogen can’t save our gas networks; it can only waste money and delay California’s work to stop climate change.

Forcing communities to use hydrogen also reduces consumer choice. People have the freedom to install electric appliances when they’re ready, using government and utility incentives. With hydrogen blending, homes and businesses would have to use a lower-quality gas whether they want it or not, safety and health risks be damned.

The California Public Utilities Commission plays a critical role protecting customers from utility investments that lock in unjustifiable rate increases. Ultimately, the Orange Cove pilot is nothing more than an expensive waste of customer money with no near-term benefit and minuscule contribution toward California’s climate efforts.

The mountain of scientific literature against hydrogen blending, lessons learned by other regulators and communities rejecting similar pilots, and the voices of Orange Cove residents should be enough to slam the door on this would-be boondoggle.

Dan Esposito is a manager in the nonpartisan think tank Energy Innovation’s fuels and chemicals program.

Activist investor Ancora Holdings Group is calling on the Warner Bros. Discovery board to consider a revised bid from Paramount Skydance and negotiate with the David Ellison-led company, or it says it will vote no on the proposed deal between Warner Bros. and Netflix.

The Cleveland-based investment management firm released a presentation Wednesday detailing why it believes Paramount’s latest offer could be a superior bid compared with the Netflix transaction.

Ancora said its stake in Warner Bros. Discovery is worth about $200 million, which would make its ownership less than 1% given the company’s $69.4-billion market cap.

Ancora cited uncertainty around the equity value and final debt allocation for the planned spinoff of Warner’s cable channels into a separate company as a factor that could change share valuation. The spinoff is still set to happen under the agreement with Netflix, as the streamer does not intend to buy the cable channels. Paramount has proposed buying the entire company.

The backing of David Ellison’s father, Oracle co-founder Larry Ellison, was a sign of the Paramount bid’s “credibility and executability,” Ancora said, adding that it had concerns about the regulatory hurdles Netflix could face.

Senators grilled Netflix Co-Chief Executive Ted Sarandos last week about potential antitrust issues related to its agreement to buy Warner Bros. Sarandos has said 80% of HBO Max subscribers in the U.S. also subscribe to Netflix and contended that a deal between the two would give the combined company 20% of the U.S. television streaming market, below the 30% threshold for a monopoly.

The investment management firm noted that Paramount is “reportedly viewed as the current administration’s ‘favored’ bidder — suggesting stronger political support,” a nod to the Ellison family’s friendly relationship with President Trump.

Trump has vacillated in his public statements on the deal. In December, he said he “would be involved” in his administration’s decision to approve any agreement, but last week, he said he “decided I shouldn’t be involved” and would leave it up to the Justice Department.

“Paramount’s latest offer has opened the door,” Ancora wrote in its presentation. “There is still a clear and immediately actionable path for the Hollywood ending that all [Warner] shareholders deserve.”

Ancora said it intends to vote no on the Netflix deal and that it also could seek to elect directors at the upcoming Warner shareholders meeting.

Warner said in a statement that its board and management team “have a proven track record of acting in the best interests of the Company and shareholders” and that they “remain resolute in our commitment to maximize value for shareholders.”

Ancora’s presentation does highlight “two primary questions as shareholders approach this deal,” said Alicia Reese, senior vice president of equity research for media and entertainment at Wedbush.

“The biggest question mark is what is Discovery Global worth?” she asked. “The second is how likely is Netflix to pass regulatory scrutiny?”

The firm’s opposition doesn’t necessarily mean the Warner board will change course, but if other significant shareholders take a similar stance, the board likely would need to “meaningfully and proactively engage further to seek more money,” said Corey Martin, a managing partner at the law firm Granderson Des Rochers.

“If I were Paramount … I would view this as a tea leaf that there might be a little bit of an opening here, to the extent we were to be aggressive,” he said. But, “if Paramount wants this company, it’s going to have to blow the Netflix bid out of the water so that there’s no question to the shareholders which bid represents the most value.”