Zeitgeist #6Saturday, September 12, 2026
Underlying Desire
At the core, this trend is about trust under uncertainty. Businesses want the speed of delegation without the fear of losing control, making a bad decision, or failing an audit. Humans have always outsourced work to tools and teams, but agentic AI forces a more primal question: can I let something act on my behalf and still remain responsible for the outcome? The desire is not just efficiency. It is the need to expand capability while preserving accountability, status, and control.
Key Evidence
According to McKinsey's 2026 global AI survey, nearly nine in ten respondents use AI in at least one business function, and 44% say AI is scaling across the enterprise, up from 38% the year before. Deloitte says about 80% of surveyed organizations still lack mature governance capabilities for agentic AI, including decision boundaries, real-time monitoring, and audit trails. McKinsey also notes the market is shifting from isolated chatbots to coding agents and agentic systems that can operate across workflows.
Why Now
Three things changed at once: adoption is broad enough to matter, agents are becoming action-oriented instead of merely generative, and governance has not kept up. McKinsey's survey shows enterprise-scale AI is no longer theoretical, while Deloitte's data shows most companies are still missing the control layer. That mismatch turns governance from a nice-to-have into budgeted software spend right now.
An enterprise AI agent governance platform that gives agents identity, scoped access, runtime guardrails, and audit trails. ([withwillow.ai](https://withwillow.ai/blog/willow-7m-seed-funding?utm_source=openai))
Outcome: $7 million seed round announced in June 2026, led by Hetz Ventures, with the platform already powering about 5,000 weekly active users at Wix. ([withwillow.ai](https://withwillow.ai/blog/willow-7m-seed-funding?utm_source=openai))
An agentic marketing operations platform that automates enterprise workflows from brief to publish while keeping governance inside the workflow. ([docs.gradial.com](https://docs.gradial.com/docs/getting-started/what-is-gradial?utm_source=openai))
Outcome: Raised $65 million in Series C in June 2026, bringing total funding to about $120 million, and it was reported as a 100-person company. ([axios.com](https://www.axios.com/2026/06/18/gradial-ai-agents-marketing?utm_source=openai))
Agent Guard is a governance and observability layer for companies deploying AI agents across internal tools and customer workflows. It would sell to mid-market and enterprise teams that need to know which agent took which action, what data it accessed, whether a human approved it, and how to replay the decision later for audit, security, or compliance. It works because most teams want to ship agents now, but their existing IAM, logging, and GRC stack was built for humans, not autonomous workflows.
Policy Flow is a workflow orchestration tool that sits between AI agents and enterprise software, enforcing permission rules, approval steps, and escalation paths before anything irreversible happens. The best customers are companies in regulated industries or complex operations, like finance, healthcare, insurance, and SaaS with large support and sales teams. It would work because businesses do not want to block automation, they want to define safe lanes for it, and that is a painful, highly specific problem that current workflow tools do not solve well.
Underlying Desire
The deeper desire here is not just legal protection. It is trust: companies want to ship AI without feeling like every launch is a gamble, and users want to know when they are interacting with synthetic content or automated decisions. Compliance software gives organizations the psychological comfort of control in a system that otherwise feels opaque, fast-moving, and politically risky.
Key Evidence
The EU AI Act's transparency obligations became enforceable on 2 August 2026, with a limited grace period until 2 December 2026 for some older systems, according to the European Commission AI Act Service Desk. The Commission's implementation timeline says most remaining AI Act provisions will not fully apply until 2 August 2028, which means the regulatory rollout is still unfolding, according to the Commission. The Commission also says its Digital Omnibus efforts are aimed at streamlining the digital rulebook, reinforcing ongoing demand for adaptable compliance tooling, according to the European Commission's digital strategy page.
Why Now
The trigger is enforcement: the AI Act moved from policy to obligation in August 2026, so companies now need working systems, not future plans. At the same time, the rules are still evolving through staged rollout and simplification efforts, which makes manual compliance especially brittle and creates a strong case for software that can update as the law changes.
AI governance software that helps regulated enterprises document, assess, and prove compliance for AI systems.
Outcome: Founded in 2018 as an ETH Zurich spin-off, with CHF 16.4M raised to date, named in Gartner's inaugural AI Governance Magic Quadrant, and customers including ETH, PwC, JobCloud, Xayn, and Armasuisse. ([modulos.ai](https://www.modulos.ai/company/?utm_source=openai))
Audit-ready privacy and AI governance software for teams that need to manage compliance, evidence, and reporting in one place.
Outcome: The company says it serves 375+ customers across 36+ countries, has a 4.7 G2 rating across 46+ verified reviews, and announced a strategic investment from Gallo Venture Capital on 5 August 2026. Its product history dates back to 2018, when it was publishing GDPR compliance software updates under the GDPR365 name. ([priviq.com](https://www.priviq.com/about-us/?utm_source=openai))
A compliance dashboard for SMBs and mid-market companies that need to know every AI system they use, who owns it, what data it touches, and which AI Act obligations apply. The product would automatically build a model inventory, generate disclosure language, track approvals, and store audit evidence in one place. It would work because most teams do not need bespoke legal advice for every workflow, they need a system of record that turns messy AI usage into something defensible and searchable.
A workflow tool for product, legal, and marketing teams that need to label synthetic content, attach provenance metadata, and keep evidence of compliance across web, app, and internal docs. It would plug into common content workflows, then auto-create the required warnings, logs, and review history whenever an AI-generated image, text block, or output is published. This works because most companies will not build this into their core systems quickly, but they still need a fast, lightweight way to avoid accidental violations.
Underlying Desire
At the core, this trend is about trust in digital money. Businesses and consumers do not actually want a token, they want speed, finality, low cost, and the feeling that their funds are safe, recoverable, and accepted by the system that already governs payroll, taxes, and audits. Stablecoin rails are becoming attractive because they promise the convenience of the internet with the credibility of regulated finance.
Key Evidence
The U.S. Treasury issued proposed rules on August 17, 2026 to implement the GENIUS Act and said the law is expected to take effect on January 18, 2027, according to the Treasury. The Federal Reserve proposed that permitted payment stablecoin issuers be treated as financial institutions under the Bank Secrecy Act, including customer identification programs, according to the Federal Reserve. The SEC proposed a startup exemption for up to $5 million over four years and a fundraising exemption up to $75 million in 12 months, according to the SEC.
Why Now
The biggest change is regulatory clarity. Treasury, the Fed, and the SEC are all moving at once, which turns stablecoins from a legal question into a product question. That means founders can finally build for distribution, compliance, and user experience instead of spending every sales cycle explaining why the business is not a compliance fire drill.
Stablecoin payments infrastructure for businesses that need to send, receive, convert, store, spend, and earn across fiat and stablecoins.
Outcome: BVNK reports 350+ team members, $39bn+ in annualized volume, and 40+ licenses globally. Its site also says it was founded in 2021. ([bvnk.com](https://www.bvnk.com/about-us?utm_source=openai))
Stablecoin and fiat payments infrastructure that lets businesses move, store, accept, issue, and spend money through APIs.
Outcome: Bridge was acquired by Stripe, which completed the acquisition on February 4, 2025. Bridge’s founder said the company had started about 2.5 years earlier, which points to a 2022 founding date. ([stripe.com](https://stripe.com/newsroom/news/stripe-completes-bridge-acquisition?utm_source=openai))
A compliance and workflow SaaS for stablecoin issuers, payment companies, and fintechs that need to launch in the U.S. without assembling a 20 person legal and ops team. It would manage license requirements, KYC rule configuration, audit trails, transaction monitoring, and regulatory reporting across federal and state regimes. This works because the market is moving into a licensed model, and companies will need software that turns fragmented regulation into an operating system instead of a spreadsheet nightmare.
A payments orchestration platform for businesses that want to accept, send, or settle via stablecoins without directly touching blockchain complexity. It would route transactions across stablecoin issuers, wallets, fiat on ramps, and treasury accounts, while giving finance teams one dashboard for reconciliation, policy controls, and settlement preferences. This would work for e commerce, payroll providers, marketplaces, and cross border B2B vendors that care about lower costs and faster settlement but do not want crypto infrastructure headaches.
Underlying Desire
At the deepest level, this trend is about control in a world where a basic utility has become uncertain again. Founders, operators, and enterprises want predictability: predictable costs, predictable uptime, predictable capacity, and predictable regulatory exposure. Electricity is turning from an invisible background input into a strategic constraint, and the desire driving this market is the same one that has always powered software adoption: people want to turn messy, variable systems into something measurable, optimizable, and trustworthy.
Key Evidence
The IEA says global power demand is rising faster in 2026 and 2027, driven by industrial growth, electrification, EVs, heat pumps, and expanding data-center capacity, according to the IEA's Electricity Mid-Year Update 2026. The IEA says data-center electricity demand grew 17% in 2025, and data centers could account for half of U.S. electricity-demand growth to 2030, according to its 2026 energy and AI analysis. The IEA also says renewables are meeting nearly 50% of the growth in data-center electricity demand, making electricity sourcing and scheduling a core software problem, not just an energy problem.
Why Now
What changed is scale. AI training and inference, plus a broader wave of electrification, pushed power demand past the point where operators can ignore it as an overhead line item. The IEA's 2026 reports show the problem is no longer hypothetical: demand growth is already accelerating, and data centers are already a material share of that growth. What also changed is the economics. Renewable supply is expanding fast enough to matter, but not evenly enough to solve procurement automatically. That gap creates a software wedge for tools that can translate grid volatility into procurement decisions, workload scheduling, and facility-level optimization.
Energy infrastructure software for data centers, with tools that help customers get power and grid connections online faster.
Outcome: Raised a $54 million Series B in June 2026, after a $20.5 million Series A in May 2024, and says it now manages 30 GW of power. ([verse.inc](https://verse.inc/newsroom/series-a?utm_source=openai))
An AI-native electricity transaction platform that matches businesses and generators while reducing energy costs.
Outcome: Tem says it serves more than 2,600 business customers in the U.K., has saved businesses over £40 million, and raised a $75 million Series B in February 2026. ([techcrunch.com](https://techcrunch.com/2026/02/09/tem-raises-75m-to-remake-electricity-markets-using-ai/?utm_source=openai))
A procurement and optimization platform for data-center operators, AI labs, and large enterprise IT teams that need to buy power intelligently instead of just paying the bill. GridPilot would ingest utility tariffs, market prices, renewable availability, and site-level load data, then recommend contract terms, hedging moves, and workload shifts that reduce total electricity spend. It works because energy has become a software-managed cost center, and most operators do not have the tooling to turn volatility into savings.
A carbon-aware workload scheduling tool for cloud-native teams that want to run jobs when electricity is cheapest and cleanest. CarbonShift would connect to Kubernetes, data pipelines, and batch workloads, then automatically shift compute across time windows or regions based on grid intensity, power price, and SLA constraints. It would sell to AI teams, SaaS companies with big batch jobs, and infrastructure-heavy enterprises that need a practical way to lower both emissions and power bills.
Underlying Desire
At the deepest level, this trend is about control. People want relief from a body or metabolism that feels resistant to effort, and they want it without living inside a clinic, a diet program, or a moral lecture. GLP-1s promise something powerful: a shortcut to stability, dignity, and a sense that health is finally manageable. For payers and providers, the same desire shows up as a need to reduce chaos, lower downstream costs, and make chronic care feel less like endless firefighting.
Key Evidence
KFF found that 18% of U.S. adults had ever used a GLP-1 and 12% were currently using one in late 2025, up six percentage points from May 2024. KFF also found that 70% of GLP-1 users said they used the drugs at least partly for a chronic condition such as diabetes or heart disease. On the payer side, KFF reported gross Medicare Part D spending on GLP-1s reached $27.5 billion in 2024, a five-fold increase from 2019.
Why Now
The category crossed from early adopter buzz into mainstream utilization, which exposes the ugly parts of reimbursement, adherence, and care coordination. At the same time, public programs are starting to pull back or tighten obesity coverage, especially in Medicaid, so patients and clinicians need help navigating inconsistent rules. That combination makes workflow software immediately valuable.
A hybrid obesity and metabolic health clinic that combines primary care, nutrition counseling, behavioral health, and GLP-1 treatment.
Outcome: Raised $25 million in October 2025, bringing total funding to more than $50 million, and says it has expanded from Massachusetts into Georgia, Illinois, Texas, and virtual care in all 50 states. The company also reports industry-leading outcomes like 93% one-year sustained weight loss and works with 1,500+ referring providers. ([fiercehealthcare.com](https://www.fiercehealthcare.com/health-tech/hlth25-cvs-health-ventures-backs-knownwells-25m-round-scale-iobesity-care-model?utm_source=openai))
A behavior-first GLP-1 and cardiometabolic care platform for employers and health plans.
Outcome: Embla says it has supported over 30,000 individuals across Europe, conducted six clinical studies, and expanded into the U.S. employer market. Its homepage also claims a fully managed program with 55% less GLP-1 use. ([joinembla.com](https://www.joinembla.com/about))
PriorPass is a benefits navigation and prior authorization platform for clinics, telehealth providers, and specialty pharmacies that prescribe GLP-1s and other high-friction chronic care drugs. It would verify coverage rules in real time, assemble the right documentation, route appeals, and track approval outcomes across payers and states. This works because coverage is fragmenting just as demand is rising, so the administrative burden is becoming a real cost center instead of a back-office annoyance.
Metabolic Loop is a longitudinal care platform for employers, payers, and obesity medicine clinics that helps patients stay on GLP-1 therapy safely and consistently. It would combine adherence reminders, side-effect check-ins, lab scheduling, nutrition prompts, symptom escalation, and clinician dashboards into one workflow. This would work because the drug is only part of the outcome, and the biggest value is in keeping patients on therapy long enough to produce real clinical and financial results.
Underlying Desire
At the core, this trend is about people trying to reduce uncertainty about their place in the world. Belonging is not just companionship, it is safety, identity, and proof that you matter to someone or some group. When life gets more fragmented, people pay for anything that reliably makes them feel seen, needed, and expected.
Key Evidence
The U.S. Surgeon General lists loneliness and social isolation as a public-health priority, according to HHS. People with strong perceptions of belonging are 2.6 times more likely to report good or excellent health than those with low belonging, according to the Surgeon General's advisory from HHS. Loneliness and isolation cost employers about $154 billion annually through stress-related absenteeism, according to the same HHS advisory.
Why Now
This became actionable when public health authorities translated loneliness into measurable costs, health outcomes, and employer pain. The category also matured because software now makes recurring coordination, matching, and engagement tracking cheap enough to productize. At the same time, traditional social platforms optimized for attention, not durable belonging. That leaves a gap for tools that create actual participation, accountability, and trusted groups rather than endless feeds.
A subscription app that matches strangers into small in-person dinners and other social events to help people make friends offline.
Outcome: Timeleft says it has seated more than 3 million guests at curated tables worldwide, and its site also cites 1.5 million events and 200+ cities across 52 countries. ([timeleft.com](https://timeleft.com/blog/dinner-with-strangers/?utm_source=openai))
A friendship and events app that helps Gen Z users discover real-world activities and meet people with similar interests.
Outcome: TechCrunch reported in September 2025 that Clyx had 50,000 active users buying tickets, more than 200,000 users browsing events, and had raised a $14 million Series A. ([techcrunch.com](https://techcrunch.com/2025/09/10/this-gen-z-founded-app-just-raised-14m-to-take-on-the-loneliness-episdemic/?utm_source=openai))
Belong Loop is a SaaS platform for employers, alumni networks, and membership organizations that turns community into a managed retention system. It helps admins create recurring small groups, auto-match people into cohorts based on interests or goals, and track participation, follow-up, and engagement over time. It would work because most organizations already pay for tools to attract people, but they have almost nothing built for keeping them connected once they arrive.
Circle Match is a local social matching tool for people who want friends, workout partners, parenting support, language practice, or hobby groups in their area. It would use lightweight profiles, intent-based matching, and recurring event prompts to turn one-off introductions into repeated interactions. This works because the hardest part of local socializing is not discovery, it is getting the first three interactions to happen consistently.
Underlying Desire
This trend is really about control without chaos. Leaders want reassurance that teams are working, managers want clarity without micromanaging, and employees want autonomy without being treated like suspects. Underneath the office attendance debate is a very human need for trust, fairness, and legibility: people want the rules to be visible, consistent, and enforceable, especially when work has become distributed and hard to observe.
Key Evidence
According to a 2025 Census Bureau working paper, 70% of firms do not track employee days in the office and 75% do not monitor employees when they work from home. The same Census research found that hybrid workers faced more return-to-office mandates than fully onsite or fully remote workers in early 2025. The Census Bureau also updated its Business Trends and Outlook Survey data in July 2026 to provide timely insight into business conditions and work-from-home measurement.
Why Now
The shift to hybrid work is no longer experimental, but the management stack never caught up. At the same time, return-to-office rules are becoming more common and more uneven, which creates a need for software that can interpret policy and measure compliance. The Census Bureau’s 2026 BTOS update signals that measurement of work-from-home is now a mainstream business data problem.
Butlr sells privacy-first occupancy sensors and workplace analytics that show how offices, floors, and rooms are actually used.
Outcome: Founded in 2019, Butlr says it has raised $68 million total, closed a $38 million Series B in August 2024, and serves customers including Qualcomm, Verizon, Netflix, Microsoft, and CBRE. ([butlr.com](https://www.butlr.com/news/butlr-secures-38m-in-series-b-funding?utm_source=openai))
ActivTrak is a work intelligence platform that measures how work happens across people, processes, and technology.
Outcome: ActivTrak says it surpassed 1 million users in 2025, grew revenue 26% that year, and received a strategic investment from Francisco Partners. Its site also says it is trusted by more than 9,500 organizations worldwide. ([activtrak.com](https://www.activtrak.com/news/press-release-momentum-2026/?utm_source=openai))
A lightweight manager analytics tool that turns messy hybrid work signals into team-level insights. Instead of surveillance, it would show patterns like collaboration overlap, office utilization, missed anchor days, and workload balance, then surface simple prompts for managers to act on. This would appeal to people leaders who are being asked to enforce hybrid norms without the data to explain what is actually happening.
A policy and attendance operating system for mid-market companies that need to track who is supposed to be where, when exceptions happen, and whether teams are complying with RTO rules. The product would let HR and ops teams define attendance policies by role, location, and team, then automatically collect check-ins, badge data, calendar context, and manager approvals into one dashboard. It would work because companies are already implementing hybrid policies, but most lack a clean system to measure them, which creates confusion for managers and employees alike.
Lev turns an idea into the things a company actually runs on: positioning, market and competitive research, a pitch deck, and a go-to-market plan.
Start building for free