CMOs are reallocating budgets to AI-native solutions, pushing agencies to run on AI, building in-house agents, and upskilling their teams. The function gets rebuilt around AI.
To pressure-test what we’re seeing, @Mayfield surveyed 50+ CMOs on how AI is reshaping the marketing org.
Key data:
72% of CMOs are actively reallocating budget to AI-native solutions
53% are increasing their overall marketing budget in 2026
74% want AI to own demand generation end-to-end
96% of CMOs said productivity and output gains are the only things that will make them switch
For founders building for CMOs, follow this playbook:
Skip the pitch. Ship the trial. Sell to Marketing Ops first.
Own the workflow, not just a step in it
Build trust into the product, not just the pitch
Price on usage or outcomes
Move now: budget is shifting, and the window is open.
This week, the enterprise AI market kept moving toward the same playbook CMOs are demanding: own the workflow, not just a step in it. OpenAI launched the OpenAI Deployment Company to embed AI directly into customer operations, Microsoft expanded Copilot Studio with agent governance and intelligent workflows, and Anthropic launched Claude for Small Business with connectors and ready-to-run workflows. The winning products will integrate with existing systems, earn trust, and deliver measurable productivity gains.
Full Weekend Edition below. 👇
Signals Shaping the Future of AI:
Infrastructure
Private equity is accelerating large-scale AI data-center buildouts. Blackstone, Apollo, KKR, and Blue Owl are pushing capital into infrastructure platforms as AI demand turns data centers into a long-duration asset class. Click here
Amazon is changing how it builds AWS data centers to future-proof them for the AI era. The internal “Titus” initiative aims to accelerate construction, expand liquid-cooling deployments, increase site capacity, reduce stranded power, and prepare facilities for higher-density Nvidia systems like GB200 and future Vera Rubin-era racks. Click here
AI data-center power procurement is shifting to behind-the-meter, GW-scale “energy parks” sold as a service. Hitachi and X LABS’ collaboration signals a move to bundled, financeable power + site solutions as a core constraint for AI buildouts. Click here
SoftBank launches a gigawatt-hour-scale battery business in Japan aimed at AI-era power constraints. The effort links energy storage supply chains directly to data center expansion, grid stability, and national AI infrastructure competitiveness. Click here
Enterprise
OpenAI launches the OpenAI Deployment Company, formalizing a forward-deployed implementation model. The structure (including a Tomoro acquisition) signals that delivery capacity, workflow integration, and SI-style execution are becoming competitive differentiators for frontier AI adoption. Click here
Microsoft expands Copilot Studio with agent governance, intelligent workflows, and connected app experiences. The release points to “control plane” expectations for enterprise agents, including centralized management, visibility, and safer automation. Click here
SAP unveils its “Autonomous Enterprise” platform with new AI agents and workflows. The company introduced a unified AI stack, domain-specific assistants, and expanded partnerships with Anthropic, AWS, Google, Microsoft, NVIDIA, and Palantir to automate enterprise operations. Click here
Anthropic reportedly surpassed OpenAI in business AI adoption for the first time. Ramp’s AI Index showed Anthropic at 34.4% enterprise adoption in April versus OpenAI at 32.3%, driven in part by strong uptake of Claude Code. Click here
Goldman Sachs says AI agents will automate parts of its “human assembly line.” The bank plans to deploy digital agents across operations and workflows while using AI to increase productivity and scale internal systems. Click here
Capital Flows
AI chipmaker Cerebras surges in public market debut. Shares jumped 68% after the company’s IPO, highlighting strong investor demand for AI infrastructure firms as OpenAI, Anthropic, and SpaceX prepare potential public offerings. Click here
Isomorphic Labs raises $2.1 billion in Series B round to scale AI-native drug discovery into real pipelines. The round focuses late-stage capital on AI life sciences platforms that can transition from model-driven discovery into owned and partnered therapeutic assets. Click here
Exaforce raises $125 million to build agentic security operations. The financing signals investor demand for AI-native SecOps platforms where autonomous investigation and machine-speed response become baseline requirements. Click here
Mind Robotics raises another $400 million for AI-powered industrial robots. The Rivian-backed startup has now raised more than $1 billion to develop robotic systems for manufacturing and industrial automation. Click here
Cowboy Space raises $275 million to build space-based data center capacity. The financing underscores how compute scarcity and infrastructure constraints are driving non-traditional approaches to scaling data infrastructure. Click here
Research
“Senses Wide Shut” identifies a representation-action gap in omnimodal LLMs. The paper advances the evaluation of multimodal agent failure modes, emphasizing the difficulty of translating rich perception into grounded, reliable actions. Click here
Researchers publish “Embodied AI in Action,” a paper that maps deployment, governance, and reliability challenges in real-world systems. The work frames the practical barriers to moving embodied agents from demos to sustained operational environments. Click here
OpenAI shares findings from its Parameter Golf machine learning challenge. The competition drew more than 2,000 submissions and showed how AI coding agents are accelerating experimentation, lowering barriers to participation, and reshaping collaborative research and model optimization workflows. Click here
Thinking Machines demonstrates AI interaction models designed for real-time responsiveness. The reporting underscores continued momentum toward lower-latency, more interactive assistant experiences as a product and capability frontier. Click here
Policy
The U.S. clears H200 chip sales to several major Chinese firms, including Alibaba and Tencent. Despite approvals, deliveries remain stalled as Washington and Beijing continue to tighten oversight of advanced AI infrastructure and supply-chain security. Click here
AI executive action is reportedly stalled by internal infighting. The delay suggests near-term AI governance may remain fragmented as competing factions debate regulatory posture and authority. Click here
xAI’s Colossus 2 power buildout is becoming a flashpoint for AI data-center permitting. The company reportedly added 19 more gas turbines in Mississippi, bringing the site to 46 turbines as it faces a Clean Air Act lawsuit from the NAACP and environmental groups. Click here
Colorado lawmakers pass a rewrite of the state’s AI law, shifting toward a more business-friendly compliance model. The update signals how state-level AI governance is being reshaped under implementation pressure ahead of broad federal rules. Click here
Global AI Strategy
The U.S. and China will begin discussions on AI safety and model safeguards. Treasury Secretary Scott Bessent said the talks will focus on protocols to prevent advanced AI systems from falling into the hands of nonstate actors. Click here
Canada funds 44 projects through its AI Compute Access Fund, turning compute policy into capacity allocation. The program operationalizes subsidized compute as an industrial-policy lever to accelerate adoption across sectors. Click here
Red Hat and Core42 launch a sovereign AI infrastructure collaboration for UAE- regulated and public-sector workloads. The partnership positions sovereign AI as a hybrid-cloud operating model combining orchestration, GPU optimization, and jurisdiction-specific governance. Click here
SoftBank’s grid-scale battery push reinforces the link between national energy strategy and AI compute expansion. Battery supply and storage capacity are emerging as integral to reliable, scalable AI infrastructure planning. Click here
Talent Signals
Each week, we spotlight key roles tied to the themes shaping this week’s AI headlines, connecting talent to the companies driving the news.
@Typeface develops enterprise AI software for marketing and content teams, helping brands generate and manage personalized creative assets at scale. As CMOs push AI deeper into campaign workflows and creative operations, Typeface is expanding across product, engineering, and customer teams. Open roles are listed on its careers page. Click here
@PhyloBio builds AI systems for biological research and discovery, applying machine learning to understand complex biological data and accelerate scientific workflows. As AI becomes more embedded in biotech and life sciences, Phylo Bio is expanding across research, engineering, and computational biology roles. Open roles are listed on its careers page. Click here
@Multiply is an AI-native media agency for B2B companies, combining expert operators with Blue, its AI agent for ad creation and performance optimization. As marketing teams shift budget toward AI-powered GTM workflows, Multiply is helping brands run more efficient, always-on media programs. Open roles are listed on its careers page. Click here
You can see all the opportunities at Mayfield-backed AI companies here.
Social Signals
The most important conversations in AI are unfolding across social media, where top voices are shaping the next wave of signals and strategy. Here are some of the top social signals and their takes from the past week.
Andrew Ng (Click here) — “There will be no AI jobpocalypse. The story that AI will lead to massive unemployment is stoking unnecessary fear. Software engineering is the sector most affected by AI tools, yet hiring remains strong. The trends point to net job creation being vastly greater than job destruction.” In a widely shared post with more than 10,000 reactions and 1,300 reposts, Ng pushes back on narratives of mass AI-driven unemployment, arguing that frontier labs, software vendors, and companies all have incentives to amplify disruption stories. He frames AI less as a collapse of work and more as a transition that reshapes workflows, creates new engineering demand, and expands productivity across the economy.
François Chollet (Click here) — “The quantity of code that devs ship has roughly 10xed. But net developer productivity, value created per unit of time, is only up by a bit, if at all. A bigger part is that the new code is creating problems of its own.” Chollet argues that AI coding tools are dramatically increasing output without necessarily increasing real business value, as larger codebases introduce more complexity, maintenance burden, and failure surface area. The post highlights a growing tension between code generation speed and the harder problem of building reliable, high-quality systems.
Arvind Jain (Click here) — “The center of gravity is shifting from the model layer to the operating layer around the model. Inside real companies, the primary bottleneck is rarely raw model capability. It’s getting AI to operate reliably across fragmented data, legacy systems, permissions, and workflows shaped by tacit knowledge.” Jain argues that the real challenge in enterprise AI is no longer the model itself, but the infrastructure and operating layer required to securely integrate, govern, and deploy AI across complex organizations.
