Meta's investment in CRED last week wasn't just about buying roughly a fifth of one of India's most valuable fintech companies. It was also about bringing CRED founder Kunal Shah in to lead WhatsApp, succeeding Will Cathcart after seven years.
It's a revealing move. WhatsApp boasts more than 3 billion users globally and over 500 million in India, yet it has barely dented the country's UPI ecosystem, accounting for roughly 0.65% of payment volume. After spending six years and $6.6 billion trying to solve that challenge through product, Meta appears to have reached a different conclusion. Sometimes the bottleneck isn't the product. It's the operator.
On June 22, Meta expanded Instagram to Samsung TVs in the US, bringing the app to one of the country's largest connected TV platforms. The app was already available on Amazon Fire TV and Google TV, meaning Instagram now has a presence across most major connected TV ecosystems in the US. Instagram is quietly evolving beyond the smartphone.
The same update introduces interest based channels, phone to TV casting, and, perhaps most tellingly, support for episodic series and Live on TV. Taken together, those features suggest Meta is designing Instagram for longer viewing sessions, not just quick bursts of attention.
The important shift isn't the device. It's the format. Reels were built for moments. Episodes and live programming are built to hold an audience. That opens the door to the kind of longer viewing sessions, premium advertising inventory, and creator programming that have traditionally belonged to television and streaming ecosystems.
The broader takeaway, as PYMNTS sees it, is that social video and streaming are beginning to compete much more directly. They're no longer fighting only for attention on your phone. They're competing for the time you spend in front of the TV as well. Meta wants Instagram to be part of that experience.
At Cannes on June 22, Reddit unveiled a suite of AI focused advertising products alongside new research into how people are shopping in the age of generative AI. One finding stood out: nearly half of shoppers say they turn to Reddit to verify AI generated recommendations before making a purchase.
That puts Reddit in an interesting position. As AI increasingly generates answers, Reddit is becoming the place people visit to sense check them. In other words, the machine may write the recommendation, but Reddit is positioning itself as the human second opinion.
On June 24, Deezer launched Remix Lab in France, giving fans a way to remix licensed songs with artist consent and per stream royalties. It's a different approach from the AI generated music tools being developed by platforms like Spotify and YouTube, where the emphasis has been on creating new content rather than reworking existing catalogues.
The timing is notable. Deezer says nearly 75,000 fully AI generated tracks are uploaded to the platform every day, accounting for about 44% of all new uploads. Rather than leaning further into generative AI, Deezer is betting that licensed creativity, backed by artists and rights holders, will prove more sustainable.
Meta used Cannes to double down on its AI advertising ambitions. On June 23, the company introduced Brand Memory, a tool that generates ad creative using a brand's own campaign history and performance data. It also repeated its claim that advertisers earn an average return of $4.13 for every $1 spent.
One detail drew particular attention: the AI creative features are enabled by default, meaning advertisers must actively opt out if they don't want to use them. For Meta, AI generated creativity is becoming the default, not the exception.
MSNBC's plans to expand its weekend video podcasts point to a broader shift in media. As veteran anchor Alex Witt prepares to step away, the network is investing further in a format that increasingly looks less like traditional podcasting and more like television delivered across multiple platforms.
For years, the biggest constraints on the attention economy were commercial. Platforms compete to attract more users, hold their attention for longer, and convert that attention into advertising and subscription revenue. Increasingly, the limiting factor is regulation.
Britain is the latest example. A House of Commons research briefing outlines proposed legislation that would prevent under-16s from accessing major social media platforms. The focus is on addictive platform features, algorithmically recommended content, and contact from strangers. If passed, the proposal would make age assurance a requirement rather than an option, reducing the number of users platforms can legally reach.
The UK is not moving in isolation. Australia has already introduced similar restrictions, and the first signs of how those rules are affecting young people's online behaviour are beginning to emerge. That gives policymakers elsewhere a real-world case study rather than a theoretical debate.
The significance extends well beyond child safety. Younger users are where many platforms establish long-term habits, build future customer value, and grow the audiences that advertisers and investors expect. If access to that cohort becomes increasingly restricted by law rather than by product design, the economics of audience growth begin to change, especially in markets that often influence regulatory thinking elsewhere.
The rails. OpenAI made its intentions a little clearer at its first appearance at Cannes on June 22. Speaking to brands and agencies, Chief Revenue Officer Denise Dresser said the company is "clearly in the advertising business now" and described a shift "from the attention economy to the intelligence economy."
The comments suggest OpenAI is no longer positioning itself as just an AI company. It's also making a case to advertisers that the next opportunity may lie in influencing decisions, not simply capturing attention.
The flow. OpenAI says ChatGPT now reaches close to a billion weekly users, and around one in five queries carries commercial intent. That means a growing number of people are arriving with a purchase already on their mind.
Sponsored placements, introduced in February beneath responses, give the company a way to participate in those moments.
What it means is that OpenAI is beginning to look like a new distribution channel for advertisers, even though many of the tools marketers rely on to measure performance are still catching up. The audience is already substantial, but the advertising ecosystem remains in its early stages.
Even so, the company's ambitions are hard to ignore. OpenAI has reportedly told investors it expects to grow from $2.5 billion in revenue this year to $100 billion by 2030, a figure that would put it within striking distance of today's largest advertising businesses.
Perhaps the more important shift is conceptual. For years, digital advertising has been built around capturing attention. OpenAI is making the case that the more valuable opportunity is reaching people when they're actively trying to decide what to do or what to buy.
By the closing bell on Friday, June 26, the AI trade had begun to split. Mega-cap AI infrastructure and memory-chip names finished lower for a fifth consecutive session as investors questioned whether 2026 hyperscaler capital expenditure, now expected to exceed $450 billion across the largest platforms, can be monetised quickly enough to justify the build-out. Sentiment weakened further after reports that OpenAI could push its IPO into next year.
The buying went elsewhere. Netflix, Spotify, Tencent Music, and SiriusXM, all trading near 52-week lows after a difficult first half, attracted fresh demand as investors rotated into businesses with established attention monetisation models through subscriptions and advertising against owned audiences, rather than companies whose valuations remain tied to AI infrastructure spending.
The macro backdrop reinforced the move. A hotter-than-expected PCE inflation reading above 4% pushed implied year-end rate hike odds towards 90%, raising the discount rate applied to long-duration growth assets. Alphabet and other AI distribution and advertising names remained under pressure, while investors showed a greater willingness to own platforms generating cash flows from existing consumer attention rather than future AI deployment.
The streaming rebound (NFLX): Netflix rose 5.5% from multi-month lows after a difficult stretch that included two failed acquisition attempts: Warner Bros. Discovery, which ultimately went to Paramount Skydance, and Roku, where Fox emerged as the winning bidder. The move reflected renewed interest in one of the market's largest direct-to-consumer media businesses, with more than 325 million paying subscribers and an advertising business expected to approach $3 billion in 2026. Investors appeared more willing to pay for proven audience monetisation than continued AI infrastructure spending.
Audio holds up (SPOT, TME, SIRI): Audio was one of Friday's stronger segments. Spotify gained 4.26% to $460.02 as investors continued to focus on its planned super-premium tier and AI-powered discovery features. Tencent Music added 2.82% to $8.39, supported by its leadership in China's digital music market and its integration with social and livestreaming products. SiriusXM also moved higher on speculation surrounding marquee talent renewals. Across all three, the common thread was straightforward: subscription revenue, advertising income, and audiences that already spend meaningful time on the platform.
Amazon's position (AMZN): Amazon added roughly 2%. The move was less about retail and more about its consumer media assets. Twitch remains one of the largest live-streaming platforms globally, while Audible and Alexa continue to anchor Amazon's position in audio. As investors differentiated between AI infrastructure and businesses already monetising consumer attention, Amazon benefited from having meaningful exposure to both.
Alphabet (GOOGL): Alphabet declined about 2%, making it one of the session's weaker attention-platform names. The investment debate remains centred on Search and YouTube, two of the world's largest content discovery and advertising businesses. Investors continue to weigh whether generative AI will expand Google's ecosystem or gradually erode the search-and-ad model that has driven its economics for more than two decades. A fourth consecutive weekly decline suggests those questions remain unresolved, particularly against a backdrop of elevated AI capital expenditure and continued competition for talent.
Snap, Reddit, and Pinterest (SNAP, RDDT, PINS): Higher-multiple, advertising-led platforms underperformed the broader rebound. With PCE inflation above 4% and market pricing close to a 90% probability of another rate hike by year-end, higher discount rates continued to pressure businesses whose valuations depend more heavily on future advertising cash flows. Snap's Spotlight, Reddit's interest-driven communities, and Pinterest's visual discovery model all sit within that group.
Bilibili (BILI): Bilibili weakened alongside Chinese ADRs and broader softness across live-streaming and creator-led video platforms. While advertising growth and margins have improved, the company remains in the earlier stages of monetising its audience. In a market rotating away from higher-beta growth assets, investors appeared less willing to pay for future engagement before it translates into more consistent cash generation.
