Charter Communications last week introduced Latis, a new household intelligence platform developed by its Intelligence Ventures division.
According to the company, Latis uses household-level signals — built on first-party data from Spectrum’s 30 million U.S. households and 500 million connected devices — to help organizations understand changes in consumer behavior and support AI-driven decisioning, automation and personalization.
John Lee, head of Intelligence Ventures at Charter, wrote on LinkedIn yesterday:
“Most data products are built to describe who consumers are or what they did. Latis is designed to help organizations understand where households are, how they’re changing, and what they’re likely to do next.
By transforming real-world household signals into continuously updated intelligence, Latis helps bridge the gap between historical data and real-time decision-making.”
Read more. (June 21)
Read: Spectrum Introduces Latis Household Intelligence Platform for the AI Era (June 18) – Charter Communications
Lee is the former Chief Data Officer at NBCUniversal.
From tipsheet: As AI makes analysis and decisioning more accessible, companies are increasingly positioning proprietary signals as a source of competitive advantage. Charter’s bet is that continuously updated household intelligence will become more valuable than static audience data.
LLMs & CHATBOTS
Developments
- A startup called Subquadratic claims it broke through a bottleneck that’s holding back LLMs (June 19) – MIT Technology Review
- Improving health intelligence in ChatGPT (June 18) – OpenAI
- Will China’s AI-powered agribots repeat its EV success story? (June 21) – South China Morning Post
COMMERCE MEDIA
Cannes: Amazon positions for agentic commerce
Amazon’s Cannes Lions programming this week includes a discussion later today titled, “Psychology of a shopper and the rise of agentic commerce,” featuring Amazon Ads VP of Devices, Content & Ads Charlotte Maines and WPP Chief Innovation Officer Elav Horwitz.
The session description asks how brands should adapt as conversational AI increasingly interprets consumer intent and influences purchase decisions.
Separately, Amazon Ads engineering VP Kelly MacLean said on LinkedIn that the company is using AI to “collapse complexity” for marketers and help them manage campaigns without needing to become builders. Read more. (June 21)
From tipsheet: If agentic advertising is one side of the equation, agentic commerce may be the other. Amazon appears to see AI assistants becoming an increasingly important layer between consumers and purchase decisions — a role it is also pursuing through Alexa for Shopping.
TECH
Cannes: Nvidia talks agentic marketing infra
In a company blog post, chipmaker Nvidia highlighted a series of partnerships across advertising and marketing ahead of this week’s Cannes Lions ad festival, including collaborations with WPP, Monks, Interpublic, Accenture Song, Adobe, Getty Images, Salesforce and ServiceNow.
The company said marketers are using Nvidia AI technologies to power content creation, audience insights, campaign optimization and autonomous marketing workflows.
Nvidia pointed to Higgsfield, a gen AI video and image platform, as an example of how its technology is being used across the marketing workflow:
“As part of the collaboration, NVIDIA Agent Toolkit software, including NVIDIA Nemotron open models, powers specialized subagents within the Higgsfield Supercomputer, running continuously inside every campaign. NemoClaw and OpenShell are being integrated to provide the enterprise trust layer.
The result: the full marketing lifecycle, from ideation and creative production through posting, performance analysis and optimization, is available in a single interface. Marketing campaigns for nearly 400 of the Fortune 500 companies are created on the platform.”
Read: “Nvidia AI Is Fueling a New Era of Marketing and Advertising” (June 19) – Nvidia blog
More: See all the Nvidia ad tech-related events at Cannes – Jamie Allan, Director, AI for Sports, AdTech & Enterprise Media, Nvidia on LinkedIn
From tipsheet: Much of the early focus in generative AI centered on creating content and creative. Nvidia’s examples point to a different destination: marketing systems that can manage increasingly large portions of the workflow, from ideation through optimization.
MARKETING
Opinion: The economics of owning the signal
In a recent essay, Henry Innis, co-founder of media measurement firm Mutinex, argues that AI may fundamentally change who captures value in media buying. For years, advertisers have relied on platform-controlled optimization and measurement systems, effectively renting intelligence from the companies selling the media.
As agentic buying, advertiser-owned measurement and direct transaction protocols emerge, Innis argues that advantage may shift toward buyers that own their own signals, models and decisioning.
He writes:
“Put the two cost curves together. The buyer can now act on every decision (volume cost gone) and reach supply directly (transaction cost gone). The only thing missing is the thing the buyer should have been bringing all along: a signal of their own.
So assemble the whole machine: buy-side signal you own, agentic execution that can act on every decision, and direct protocols that reach supply without the middleman tax. That combination doesn’t make the existing market more efficient. It rewrites who captures the value.”
Read: “Can AI Re-Make the Economics of Buying Media?” (June 21) – Henry Innis on his “Marketing Economics” Substack
From tipsheet: The piece echoes a theme emerging across ad tech: intelligence may be shifting from the platform to the buyer. As AI lowers the cost of execution, competitive advantage may increasingly come from proprietary signals, measurement and optimization models.
PLATFORMS
Q&A: WBD on its new agentic platform
Last week, WBD (Warner Bros. Discovery) announced an agentic AI-powered advertising platform built with AWS. The company said the system uses autonomous agents across audience planning, forecasting, optimization and measurement while unifying workflows across linear, streaming and digital channels.
A planned rollout is ongoing throughout 2026.
Read: Warner Bros. Discovery Announces Agentic AI-Powered Advertising Technology Built with AWS – Warner Bros. Discovery (June 17)
Dr. Nage Sethu, SVP, Technology – Converged Advertising and Linear Systems, Warner Bros. Discovery, answered a selection of follow-up questions from tipsheet via email.
tipsheet: “The announcement describes autonomous agents for planning, forecasting, optimization, and measurement. Can you say more about what decisions agents can make and what need human approval? Curious about the dividing line today.”
Dr. Nage Sethu: “The agents are being designed to automate analysis, recommendations, forecasting, pacing, and optimization decisions that traditionally require significant manual effort. They can evaluate thousands of variables in real time and recommend the best allocation of budget and inventory.
However, strategic decisions remain firmly in human hands (human in the loop). Advertisers and sales teams determine campaign objectives, budgets, brand requirements, and business priorities. The agents provide intelligence and execution support, while humans retain oversight and final approval of key business decisions.”
Additional questions include:
- WBD is rebuilding its advertising infrastructure with AWS agentic AI. What problem was WBD trying to solve?
- The platform promises ongoing automated campaign optimization. What is the biggest decision that advertisers will no longer have to make themselves?
- WBD says the platform unifies buying across linear and digital inventory. How would you say that changes the advertiser experience?
- Regarding the platform’s flexible targeting across brands, audiences, linear, and digital inventory, how does AI determine inventory allocation recommendations?
Read: Q&A: WBD building ‘agentic platform that unifies’ (June 21) – tipsheet
PLATFORMS
Seufert: On AI and Meta beyond the walled garden
Mobile Dev Memo analyst Eric Seufert argues that Meta is unlikely to systematically target non-IDFA traffic through Facebook Audience Network (FAN) on iOS, despite ongoing speculation within the mobile app ecosystem.
He notes that a brief increase in FAN activity earlier this year was likely the result of a product change rather than a broader shift in targeting strategy.
Seufert writes:
“As Meta’s advertising business becomes increasingly differentiated by proprietary AI infrastructure, it captures far more of the economic value from applying those capabilities to the inventory it owns directly than to third-party inventory served in an auction environment that prevents structural disadvantages to it.”
While some observers believe Meta could use its AI capabilities to compete more aggressively for non-IDFA mobile gaming traffic, Seufert contends that the opportunity may be too small relative to the growth potential of Meta’s owned-and-operated properties.
Read: “Would Meta reprioritize non-IDFA traffic in FAN?” (June 19) – Eric Seufert on Mobile Dev Memo (subscription)
AGENTS
AdCP protocol adds verification, measurement
Ad Context Protocol (ADCP) released version 3.1, with “signed” brand verification, vendor-attested measurement, product-feed mirroring, standardized creative formats and stronger compliance controls.
Scope3 CEO Brian O’Kelley, a member of the consortium that founded AdCP, said last week:
“AdCP 3.1 is now GA….
Headline additions:
- Release-precision version negotiation and version-scoped badges
- Distributed brand.json and signed brand verification responses
- Canonical creative formats and list_transformers
- Wholesale product/signal feed mirroring with conditional fetch + webhooks
- Product-scoped signal targeting
- Media-buy impairments, webhook activity, action discovery, proposal cleanup
- Vendor-attested measurement, delivery/usage finality, and stronger runtime/idempotency rules
- Expanded 3.1 compliance/storyboard coverage”
Read more on LinkedIn. (June 18)
More: What’s New in AdCP 3.1 – Ad Context Protocol website
From tipsheet: These early protocol releases increasingly resemble the infrastructure that emerged in programmatic advertising as ecosystems matured and, eventually, commercial activity scaled.
AGENCIES
Publicis’ Sadoun on LiveRamp but not LiveRamp
Arthur Sadoun, CEO, Publicis Groupe last week:
“Most clients want change, but AI is accelerating faster than many structures can adapt, pushing existing systems and ways of working to the brink. That’s where our role as a connective tissue comes in.
We help clients move beyond silos and connect capabilities across the full customer journey, so AI is driving more consistent decisioning. At the same time, we focus on data and identity, because agentic AI is only as powerful as the data behind it. That means helping clients build a clear, unified view of the consumer, in a privacy-first environment.
We are also redesigning workflows…”
Read: “Arthur Sadoun: ‘AI on its own is no longer a competitive advantage’“ (June 18) – Creative Salon
From tipsheet: Sadoun argues that AI is only as powerful as the data and workflows behind it. Publicis’ planned acquisition of LiveRamp reflects a broader bet that competitive advantage will come from connecting identity and data with decisioning across the enterprise, rather than from AI alone.
DATA
Hightouch CEO explains LiveRamp interest
Axios reported early last week that now-agentic CDP firm Hightouch had tried to purchase part of the assets of LiveRamp after agency holding company Publicis Groupe agreed to acquire the data collaboration company.
On Friday, Hightouch CEO Tejas Manohar admitted the attempt:
“Did Hightouch make an offer to buy a portion of LiveRamp? Yes, we did.
‘Wait, why would you do that!?’ — the internet.
A few reasons:
- We already sell a competing data onboarding and measurement product, MatchBooster, and tend to win head-to-head against LiveRamp on performance, usability, speed, and security.
- Since the Publicis-LiveRamp news broke, brands and agencies have come to us interested in neutral alternatives. Fair or not, it’s real market sentiment.
- We’d love the opportunity to bring our more modern technology to the large customer base LiveRamp has built over 10+ years.
We believe this scenario would be a win-win-win: LiveRamp gets its Publicis acquisition, Publicis gets many of LiveRamp’s most valuable assets plus software DNA to drive transformation in the age of AI, and Hightouch accelerates its market share in adtech and data onboarding…”
Read more on LinkedIn. (June 19)
Earlier in the week: “So they think Publicis wants the onboarding pipes but not the identity spine. Interesting.” (June 16) – Ari Paparo, Markecture Media on X
SEARCH
Where Search is heading
Dan Roberts, Global SVP Search, Stagwell Media Platform discussed where he thinks Search is heading:
“We’re moving into environments where discovery, evaluation and recommendation sit together; and where visibility is earned inside the flow of decision-making, not just on a results page.
From a Stagwell Search+ perspective, this is the direction of travel:
- Search as presence within AI systems, not just rankings
- Earlier influence across the decision journey
- Blending of search, commerce and recommendation into one experience“
Read more on LinkedIn. (June 19)
Related: Lenovo sets the pace for AI-led search and shopping with a 106% increase in ROAS (June 8) – Microsoft Advertising blog
From tipsheet: Search, commerce and recommendation are converging.
MORE
- Former DataXu founder Sandro Cantanzaro on his new CTV ad startup, Momentiv (June 21) – LinkedIn
- L’Oréal accelerates generative AI content engine with fresh OpenAI deal (June 19) – Digiday (subscription)
- “Accenture Song has a new tool in its measurement box—Amazon data—which it claims will help solve the problem of platforms grading their own homework.” (June 19) – Garett Sloane, senior editor of technology and AI, Ad Age on LinkedIn
- Index Exchange Must Face Wiretap Suit, Judge Rules (June 18) – MediaPost
- A Founder Ran the Numbers on AI-Generated Creative Content. The Results Were a Reality Check (June 19) – Inc.