
Work Here?
Scope3 measures and reduces carbon emissions in the media and advertising industry. It collects and analyzes emissions data across the programmatic supply chain to quantify a campaign's carbon footprint. Clients access its data and analytics to understand footprints and make reductions, with partnerships with publishers to fund carbon removal projects. Its public benefit corporation status and focus on media supply chains set it apart, aiming to help organizations meet sustainability targets and decarbonize the industry.
Industries
Data & Analytics
Enterprise Software
Social Impact
Company Size
51-200
Company Stage
Late Stage VC
Total Funding
$45M
Headquarters
New York City, New York
Founded
2022
See people who can refer or advise you
Help us improve and share your feedback! Did you find this helpful?
Total Funding
$45M
Above
Industry Average
Funded Over
2 Rounds
Health Insurance
Dental Insurance
Vision Insurance
Health Savings Account/Flexible Spending Account
Disability Insurance
401(k) Retirement Plan
401(k) Company Match
Paid Vacation
Paid Holidays
Parental Leave
Remote Work Options
Home Office Stipend
Stock Options
Digest: Google under CMA scrutiny; InMobi & Scope3 launch sell-side AI agent. In today's Digest, we cover Google coming under UK scrutiny over its use of publisher content in AI products, InMobi and Scope3 launching an autonomous media buying agent, and DeepSeek nearing a $7bn funding round. Google under CMA scrutiny. UK regulators have ordered Google to give publishers greater control over how their content is used in AI-generated search results. Under measures announced by the UK's Competition and Markets Authority (CMA), publishers must be able to control whether their material appears in AI-generated features, receive clearer information about how their content is used, and gain access to more detailed performance metrics. Google will also be required to improve attribution in AI search responses and allow publishers to opt out of content being used to fine-tune AI models without having to remove themselves from traditional search results. In response, Google said it is testing new controls that allow website owners to manage how their links and content appear in AI-powered search experiences. The company said it is also expanding the use of inline links, website previews and source attribution within AI-generated answers to encourage users to visit original publishers. The CMA said the changes are intended to strengthen publishers' bargaining power as AI transforms search behaviour, while improving transparency and trust for users. InMobi & Scope3 launch sell-side AI agent. InMobi is betting on AI agents to sell its premium commerce inventory, launching a new sell-side AI agent developed in partnership with Scope3. The tool is designed to act as an autonomous sales representative, allowing advertisers to directly access inventory from Glance and InMobi's broader ad marketplace. Through the platform, brands can reach more than 300 million active devices and engage consumers through AI-powered shopping experiences that combine product discovery, virtual try-ons and direct purchasing. According to the company, AI agents could eventually streamline media planning, negotiation and campaign optimisation, reducing reliance on traditional demand-side and supply-side platforms. InMobi argues that Glance's commerce-focused inventory is better suited to direct agent-to-agent transactions, where AI systems can negotiate value based on context and user intent rather than through conventional programmatic auctions. DeepSeek nears $7bn funding round. DeepSeek is close to completing its first external fundraising round, raising more than 50bn yuan (£5.85bn) at a valuation of just under USD$60bn (£47.4bn), according to sources familiar with the matter. The deal would represent a dramatic increase from the company's reported USD$10bn (£7.9bn) valuation in April, underscoring growing investor confidence in its AI capabilities and market potential. Market-focused investors and several Chinese technology companies are said to have pledged around 30 billion yuan (£3.24bn) toward the round, highlighting strong demand for exposure to one of China's fastest-rising AI firms amid intensifying competition in the sector.
InMobi Advertising has partnered with Scope3 to make its premium connected TV and mobile in-app inventory accessible to AI-powered buying agents through Scope3 Interchange. The integration uses the Ad Context Protocol from AgenticAdvertising.org, reducing setup time from days to minutes. The partnership gives buying agents access to approximately two billion mobile users globally and first-party behavioural data from Glance's 300 million active devices. Glance, InMobi Group's consumer platform, operates across mobile and CTV surfaces including lock screens. The agentic buying model allows premium inventory to maintain its value rather than being commoditised through traditional programmatic auctions. Pilot campaigns are expected to launch in the coming weeks, marking the first agentic buying workflow for mobile in-app supply at this scale.
AdCP & the rise of Agentic Media Buying. Scope3's Ad Context Protocol enables AI agents to negotiate, buy, and optimise advertising without traditional DSP/SSP infrastructure. With Amazon DSP, Magnite, and Index Exchange as launch partners, agentic media buying is going live. Founder, Advertising Protocols March 14, 2026 On March 13, 2026, Scope3 launched its Agentic Media Platform with Amazon DSP as the first integration partner, alongside Index Exchange, Equativ, and media.net as launch distribution partners. This was not a concept demo - it was a production deployment of agent-to-agent advertising transactions using the Ad Context Protocol (AdCP). AdCP, first proposed by Scope3 in October 2025, is an open protocol for agent-to-agent advertising transactions built on Anthropic's Model Context Protocol (MCP). Where traditional programmatic uses OpenRTB bid requests flowing between deterministic software systems, AdCP enables AI buyer and seller agents to negotiate ad placements using natural language intent, structured context, and autonomous decision-making. How AdCP differs from OpenRTB. Understanding AdCP requires understanding what it replaces - and what it does not. OpenRTB operates on a request-response model: an SSP sends a bid request describing available inventory, DSPs respond with bids. The entire exchange is structured, deterministic, and happens in under 100 milliseconds. There is no negotiation - only a single take-it-or-leave-it auction. AdCP operates on an intent-matching model. A buyer agent expresses campaign intent (target audience, objectives, budget constraints, brand safety requirements) as structured context. A seller agent interprets that intent against available inventory, audiences, and contextual signals. The agents negotiate - they can counter-propose, ask clarifying questions, and iterate toward agreement. This is not replacing real-time bidding for standard display impressions. AdCP targets the campaign setup, deal negotiation, and strategic buying layer - the work that currently requires days of emails, IOs, and manual deal configuration. The technical architecture. AdCP is built on three layers: MCP Foundation: Anthropic's Model Context Protocol provides the agent-to-tool communication layer. MCP defines how AI agents call structured tools, receive responses, and maintain context across multi-turn interactions. AdCP Schema Layer: AdCP adds advertising-specific schemas on top of MCP - structured representations of campaign briefs, inventory descriptions, audience segments, pricing models, and deal terms. Agent Runtime: The actual buyer and seller agents that use LLMs to interpret campaign goals, match them against inventory, and negotiate terms. These agents run on infrastructure provided by the buyer (DSP) and seller (SSP/publisher) respectively. Magnite's seller agent: the first SSP implementation. Magnite was the first SSP to build a seller agent for AdCP, embedding it directly into SpringServe (their CTV ad server). The seller agent was tested in December 2025 with Scope3's buyer agent, and additional pilots followed in early 2026 with LG Ad Solutions, MiQ, and Warner Bros. Discovery. In the Magnite implementation, the seller agent uses an LLM to interpret the buyer agent's campaign intent and match it against available CTV inventory in SpringServe. When both agents agree on terms, a campaign is placed - without a human manually configuring line items, setting up deals in a UI, or exchanging insertion orders via email. The Scope3 Agentic Media Platform. Scope3's March 2026 launch formalized AdCP into a production platform with two key products: Agentic Media Platform: The core product enabling buyer agents to discover and transact with seller agents across the AdCP ecosystem. Amazon DSP is the first demand-side integration, meaning Amazon's buying infrastructure can route campaigns through AdCP-enabled seller agents. Brand Standards: A companion product that encodes brand safety and suitability requirements as structured agent context - ensuring that agentic buying decisions respect advertiser brand guidelines without requiring manual inclusion/exclusion list management. AdCP vs. AAMP: complementary, not competing. A common question: does AdCP compete with IAB Tech Lab's AAMP (Agentic Ad Management Protocols)? The short answer is no - they address different layers. AAMP provides the trust and governance framework (Agent Registry, identity verification, capability declarations) and the execution control plane (ARTF for real-time agentic decisioning). AdCP provides the agent-to-agent transaction protocol - how buyer and seller agents actually communicate campaign intent and negotiate deals. In practice, an AdCP buyer agent would be registered in AAMP's Agent Registry, use ARTF for real-time execution, and use AdCP for deal negotiation. The two frameworks are complementary. What this means practically. For DSPs: AdCP integration means your platform can express campaign requirements as structured agent intent and participate in automated deal negotiation. This is most immediately valuable for CTV and premium programmatic guaranteed deals - the high-touch, high-margin transactions that currently require manual setup. For SSPs and publishers: Building a seller agent means your inventory can be discovered and transacted by AI buyer agents without manual deal configuration. Early movers (Magnite, Index Exchange, Equativ) gain a distribution advantage as more demand flows through agentic channels. For agencies: Agentic media buying does not eliminate the agency - it eliminates the manual operational work of deal setup, IO processing, and campaign trafficking. The strategic decisions (which publishers to work with, what audiences to target, what creative strategy to deploy) remain human decisions, expressed as intent that agents execute. For advertisers: Faster campaign activation, more inventory discovery (agents can evaluate thousands of seller agents simultaneously), and potentially lower operational costs as manual campaign setup is automated. The open questions. AdCP is production-ready for CTV and premium direct deals, but several open questions remain: Liability: When an AI agent makes a buying decision that violates brand safety guidelines, who is liable - the agent operator, the LLM provider, or the platform? Audit trails: How do you audit an AI agent's negotiation reasoning? AdCP transactions need the same transparency that SupplyChain objects brought to programmatic - but for AI decision-making. Interoperability: Can a buyer agent built on GPT-4 negotiate effectively with a seller agent built on Claude? Protocol-level interoperability is defined, but LLM-level reasoning compatibility is not guaranteed. Scale: AdCP currently targets high-value, low-frequency transactions (PMP deals, PG campaigns). Whether it can scale to the millions-of-decisions-per-second pace of open-exchange RTB remains to be proven. The advertising industry has spent two decades building programmatic infrastructure for software-to-software transactions. AdCP represents the beginning of the next era: agent-to-agent transactions where AI systems negotiate, buy, and optimise media on behalf of humans. The protocol is live, the partners are real, and the production deployments are running. AdCP Scope3 Agentic AI MCP Amazon DSP Magnite Media Buying Founder, Advertising Protocols Samraj Matharu is the founder of Advertising Protocols - the open reference for ad tech standards, protocols, and industry knowledge. Explore all protocols. Deep-dive into the technical specs behind the ads you see every day.
Campaign Ad Net Zero Awards 2025: winners revealed. MG OMD, ITV and Scope3 were among the winners at the fourth annual Campaign Ad Net Zero Awards, which took place at The Londoner Hotel. * Limited free articles a month * Free email bulletins * Unrestricted access to our most popular content including the School Reports, Power 100 and Best Places to Work * Essential round-ups such as Movers and Shakers, Pitch Update and the Question of the Week * Leading insights from Campaign's award-winning editors
Scope3 has promoted Tim Collier to chief commercial officer, effective immediately, as the sustainability-focused adtech firm doubles down on what it calls "agentic advertising."
Find jobs on Simplify and start your career today
Industries
Data & Analytics
Enterprise Software
Social Impact
Company Size
51-200
Company Stage
Late Stage VC
Total Funding
$45M
Headquarters
New York City, New York
Founded
2022
Find jobs on Simplify and start your career today