The digital economy is grappling with a fundamental identity crisis as AI agents proliferate. Today's top stories track the rapid development of a new trust infrastructure, with vendors racing to provide agents with their own verifiable credentials. This push for 'Know Your Agent' is being met with a parallel effort to create verifiable business identities, a critical step for securing machine-to-machine commerce.
Identity authorization network Proof and business data platform Enigma have partnered to launch 'Business Certificates,' a new solution for continuously verifying business identity. Announced on Thursday, the certificates aim to combat fraud in agentic commerce by providing a dynamic, verifiable credential for individuals and AI systems acting on a business's behalf. The offering integrates Enigma's Know Your Business (KYB) data with Proof's authorization infrastructure, addressing risks like payment fraud and vendor impersonation.
Why it matters
This partnership addresses a critical missing piece of the agentic trust puzzle: verifying the business entity an agent represents, not just the agent itself. While 'Know Your Agent' (KYA) standards are emerging, they are insufficient without a corresponding 'Know Your Business' (KYB) layer that is dynamic, not static. For B2B agentic commerce to scale, both sides of the transaction require verifiable, continuously updated credentials. This moves beyond a one-time check to a persistent state of verification, which is essential infrastructure for any founder building tools for autonomous B2B transactions.
The announcement positions this as a direct response to the rise of AI-driven fraud, where deepfakes and sophisticated impersonation tactics make traditional, static identity checks obsolete. Experts see this as a necessary evolution from point-in-time KYB to a live, queryable system of business identity, crucial for high-stakes, high-velocity agentic transactions.
A new analysis from OpenZeppelin on Wednesday formally maps the emerging infrastructure for agentic payments, contrasting crypto-native blockchain standards like the MPP protocol we've tracked against consortium frameworks from Visa and Mastercard. The report also serves as the source for the data we noted earlier this week—that $73 million was settled across 176 million blockchain transactions by agents over the past year—confirming stablecoins are dominating early volume.
Why it matters
This analysis provides a clear architectural map of the competing ecosystems for machine-to-machine commerce. For builders, it clarifies the strategic choice between open, permissionless protocols and closed, consortium-led systems. The dominance of stablecoins in early volume is a strong signal that crypto-native rails are currently better suited for the high-frequency, low-value transactions characteristic of agentic economies. The key thing to watch is which track gains institutional traction for larger-value B2B settlements.
Experts in the report suggest the two tracks are not mutually exclusive and may eventually converge, with card networks potentially using public blockchains for settlement. The analysis emphasizes that both approaches are grappling with the same core challenges: establishing verifiable agent identity, creating robust authorization frameworks, and defining liability for erroneous or fraudulent transactions.
The push to classify AI agents as a 'first-class identity'—a concept we noted gaining traction from Entrust's CIO earlier this month—has led to a major acquisition. Vercel has acquired identity and authorization startup Better Auth, explicitly framing the deal around giving agents their own verifiable credentials rather than having them borrow human developer accounts for tasks like opening pull requests.
Why it matters
This acquisition signals that agent identity is becoming a core piece of developer infrastructure, not just a security add-on. For founders, this is a critical development. It means the tools and platforms you build on will increasingly demand that your AI agents have their own credentials, impacting security posture, access controls, and auditability. The era of generic 'bot' accounts with over-privileged access is ending, forcing a more disciplined approach to managing non-human actors in the development lifecycle.
Security analysts view this as a necessary step to address the significant audit and security gaps created by agents using shared or human credentials. The move is seen as part of a broader industry shift toward treating non-human identities with the same rigor as human ones, enabling granular permissions and clear accountability for agent actions.
A Forbes analysis on Wednesday warns that enterprises face an impending 'Agentic AI bill moment,' mirroring the unexpected cost explosion seen with early cloud adoption. The piece argues that agentic AI's high token consumption, complex data generation, and rapid, decentralized deployment by business units are creating unmanageable observability costs and project delays. CIOs are urged to proactively implement strategies for visibility, compliance, and cost attribution before expenses spiral out of control.
Why it matters
This is a critical warning for any founder building or selling agentic AI solutions. Your customers' biggest adoption barrier may not be technical feasibility but financial predictability. The analysis highlights a significant market need for tools that provide cost attribution, governance, and observability for AI agents. Positioning a product to solve this emerging 'token cost' problem could be a powerful GTM angle, as enterprises will be desperate for solutions that prevent runaway spending and provide clear ROI.
Andi Mann, Chief Product and Technology Officer at Apica, who is quoted in the article, emphasizes that without proactive management, observability costs for agentic systems can quickly dwarf the cost of the AI models themselves. He suggests that, like the FinOps discipline that emerged to control cloud spend, a new 'AIOps' or 'AgentOps' discipline is needed to manage the financial and operational overhead of autonomous agents.
A new project, CONSO, is building a decentralized consumer reputation layer on the Sui blockchain, aiming to solve the problem of fragmented digital identity. According to announcements on Wednesday, the system will allow users to aggregate their activities into a portable 'Consumer Passport' with verifiable credentials. A key feature is 'Agentic Memory,' designed to provide a persistent, verifiable memory for AI systems, powered by the Walrus Protocol.
Why it matters
This project directly tackles two fundamental problems for agentic AI: lack of persistent memory and lack of verifiable reputation. By creating a user-owned, on-chain memory layer, CONSO could enable AI agents to operate with context and history across different applications, evolving beyond their current 'forgetful' state. For builders, this represents a potential infrastructure layer for creating more capable and trustworthy agents that can build reputation over time, a core component of any robust trust system.
The teams behind Sui and Walrus Protocol are positioning this as a key use case for their technology, emphasizing the need for high-throughput, low-cost blockchain infrastructure to support the millions of micro-transactions required to update an agent's memory and reputation state in real-time. Skeptics question whether users will consent to having their cross-platform activities aggregated, even in a decentralized manner.
An analysis in ITWeb on Wednesday argues that agentic commerce inverts traditional cybersecurity paradigms. The old model focused on filtering out bots; the new model must welcome legitimate AI agents while verifying their authority. This requires a fundamental shift from 'prove you're human' to 'prove you're an authorized agent acting within a defined scope.' The article details the new challenges this creates, including machine-speed fraud and novel prompt injection risks.
Why it matters
This piece crystallizes the core identity problem at the heart of the agentic economy. It's not about stopping machines; it's about discerning good machines from bad ones. For founders building in this space, this framework is essential. Your product's success will depend on its ability to manage agent identity, verify intent, and recalibrate fraud detection for non-human actors. The article correctly identifies that intent itself has become a security problem, requiring new governance and trust infrastructure.
Security leaders cited in the piece stress that traditional security tools are blind to this new threat model. They advocate for recalibrating fraud detection systems to look for anomalies in agent behavior rather than simply blocking non-human traffic. The consensus is that a robust agent identity and access management (IAM) layer is no longer a 'nice-to-have' but a prerequisite for any enterprise deploying autonomous agents.
As enterprises adopt agentic AI, they face a crisis with existing Identity and Access Management (IAM) systems, which were not built for non-human actors. A CSOonline article on Wednesday proposes a six-stage maturity model to help organizations govern AI agents. The framework progresses from basic agent inventories to advanced models featuring unique attribution, 'on-behalf-of' permissions, just-in-time credentialing, and real-time revocation capabilities.
Why it matters
This maturity model provides a much-needed strategic roadmap for enterprises struggling with agent governance. For founders selling into the enterprise, this is your new sales framework. Understanding where a potential customer sits on this maturity curve allows you to tailor your pitch to their specific pain points, whether it's basic visibility (Stage 1) or the need for runtime policy enforcement (Stage 5-6). It codifies the trust and accountability problem into a set of concrete operational steps.
The model emphasizes that achieving the highest levels of maturity requires a shift from static, role-based access control (RBAC) to dynamic, attribute-based (ABAC) and intent-based (IBAC) systems. Security analysts agree that this evolution is necessary to manage the inherent non-determinism of AI agents and ensure that their actions remain bounded and auditable.
Following the G2 data we highlighted yesterday showing that 51% of B2B buyers now start software evaluations with AI chatbots, being recommended by models like ChatGPT has become a critical GTM objective. An analysis on Wednesday explains that AI recommendations are based on ecosystem-wide credibility rather than just a brand's website, requiring consensus across trusted third-party sources like review sites and forums.
Why it matters
This marks a structural shift in B2B discovery, moving from search engine optimization (SEO) to what is being called 'Generative Engine Optimization' (GEO). For early-stage companies, the playbook is no longer just about keyword ranking; it's about manufacturing consensus. This requires a deliberate strategy of securing positive mentions and detailed descriptions across a distributed set of credible, human-curated sources that AI models are known to trust. A company's own marketing copy is now secondary to what the rest of the internet says about it.
Marketing strategists describe this as a move from 'owning the keyword' to 'owning the answer.' The focus is on providing specific, data-rich content and ensuring it is reflected in third-party analysis. Another key tactic highlighted is ensuring your company's information is present and accurately structured on platforms like Wikipedia and industry-specific wikis, as these are often weighted heavily by AI models.
A 2026 analysis published Wednesday reveals a stark trade-off in B2B sales development. While AI SDRs can deliver 10-50 times the outreach volume of human SDRs at a fraction of the cost, their meeting show rates and opportunity conversion rates are significantly lower. Consequently, a hybrid model is emerging as the most effective GTM approach, using AI for high-volume, top-of-funnel prospecting and qualification, while humans handle nuanced objection handling and relationship building.
Why it matters
This provides a crucial counterintuitive signal for founder-led sales: scaling outreach with AI is not a silver bullet. The data shows that raw volume comes at the cost of quality and conversion, and over-reliance on pure automation can damage a startup's domain reputation. The structural shift toward a hybrid model is a key playbook for early-stage companies. It suggests the optimal strategy is to use AI to generate and qualify leads, but ensure a human is looped in to manage the highest-intent conversations, balancing scalability with effectiveness.
Sales leaders caution that the hidden cost of pure AI outreach is a burned domain and a fatigued total addressable market. The consensus is that AI's best use is as a force multiplier for human reps, automating research and initial contact so they can focus on high-value interactions. This also requires a shift in metrics, moving from pure volume (emails sent) to quality (meetings booked and attended).
LinkedIn has become the most-cited domain for professional queries by major AI search engines, but a new analysis on Wednesday shows a surprising trend: 75% of these citations are sourced from the profiles of individual members, not official company pages. This suggests AI models are prioritizing demonstrated individual expertise and original, data-rich content published by employees over branded corporate content.
Why it matters
This is a major counterintuitive signal for B2B GTM strategy. It means that activating and empowering subject matter experts within your company to publish on LinkedIn is now a more effective strategy for AI-driven discovery than investing in your corporate page. For founders, the playbook is to encourage key employees to build their personal brands, share specific, educational content, and showcase original data on their profiles. Your team's collective authority is now a more powerful distribution channel than your company's own voice.
Generative Engine Optimization (GEO) experts note that AI models are designed to find and synthesize consensus from authoritative sources, and they appear to be treating individual experts with deep, specific content as more authoritative than generic marketing from a brand. This reinforces the idea that founder-led and expert-led content is becoming paramount in B2B discovery.
Singapore's OCBC Bank has launched GOLDX, a tokenized gold fund available to institutional investors, hedge funds, and asset managers on both the Ethereum and Solana public blockchains. Announced Thursday, the fund allows qualified investors to use stablecoins or fiat to acquire tokenized shares, which are delivered directly to their on-chain wallets.
Why it matters
This is a significant step in the convergence of traditional finance and public blockchains, demonstrating a major regulated institution's comfort with a multi-chain strategy for real-world asset (RWA) tokenization. Unlike many early pilots, this is a live, commercially available product for institutional clients. It signals that institutions are increasingly viewing public chains like Ethereum and Solana as viable, interoperable settlement layers for regulated financial products, a key milestone for the 'Ethereum merges into the broader economy' thesis.
Industry analysts see this as a sign of the maturing RWA ecosystem, where the debate is shifting from 'if' to 'which chain.' The choice to launch on both Ethereum and Solana highlights a pragmatic, chain-agnostic approach by institutions, who are prioritizing access and functionality over protocol maximalism.
Dinari, a platform for tokenizing US company shares, and tZERO, a regulated alternative trading system (ATS), announced a partnership on Thursday to launch a white-label solution for issuing tokenized equities. The turnkey platform is designed to lower the barrier for broker-dealers, asset managers, and fintech firms to create their own compliant tokenized stock offerings.
Why it matters
This partnership addresses a major bottleneck for the institutional adoption of tokenized real-world assets: the lack of compliant, accessible infrastructure. By bundling Dinari's tokenization tech with tZERO's regulated trading venue, they are creating a 'RWA-in-a-box' solution. This could significantly accelerate the on-chain availability of US stocks, a crucial step in bridging DeFi with traditional equity markets and building a more integrated financial system on public blockchain rails.
The move is seen as a maturation of the RWA space, shifting from bespoke, one-off projects to scalable, repeatable infrastructure. Analysts suggest that by making it easier for existing regulated entities to issue and trade tokenized assets, the partnership could unlock a new wave of institutional activity that has so far been sidelined by regulatory uncertainty and high implementation costs.
Lido's LDO token surged over 16% on Wednesday following two major integrations. The liquid staking protocol now allows direct transfers of its wrapped staked Ether (wstETH) to Robinhood, opening access for retail investors. Simultaneously, a partnership with Anchorage Digital Bank will allow institutional clients to mint and redeem wstETH within a regulated environment, further enhanced by Lido's new Web3SOC compliance certification.
Why it matters
This two-pronged expansion provides crucial new on-ramps for both retail and institutional capital into the Ethereum staking ecosystem. By integrating with regulated platforms like Robinhood and Anchorage, Lido is significantly de-risking participation and lowering the barrier to entry. This is a clear example of Ethereum's core economic activity converging with the traditional financial system, moving staked ETH from a crypto-native niche to a more accessible yield-bearing asset for a mainstream audience.
The partnership with Anchorage Digital, a federally chartered crypto bank, is seen as particularly significant for institutional capture. It provides a compliant and secure custody solution that large funds require. The Robinhood integration, meanwhile, could dramatically increase the retail user base for liquid staking, though some critics worry it centralizes access through a mainstream brokerage.
Building on the Harvard/INSEAD study we tracked earlier this week showing AI-native startups operate with significantly fewer workers, a new analysis details the specific playbook these founders use to hit $1M in ARR before seeking institutional capital. The approach emphasizes using AI as an 'operating system' to achieve extreme capital efficiency, enabling small 'micro-unicorn' teams to generate unprecedented revenue per employee.
Why it matters
This trend fundamentally alters the founder journey, especially for companies in the $0-10M stage. The ability to achieve significant traction and 'proof before pitch' with a tiny, AI-augmented team shifts the power dynamic in fundraising. It challenges conventional wisdom on hiring timelines and team composition, suggesting that the first key hires might be systems and automations, not people. For founders, this means success is increasingly defined by the ability to architect an efficient, AI-native operating model.
Venture capitalists are taking note, with some funds now specifically targeting these highly capital-efficient, bootstrapped-to-scale companies. However, other investors express concern that while AI makes it easier to build, it doesn't solve the hard problem of distribution, and many of these lean startups may struggle to cross the chasm from initial product to durable market presence.
Expanding on the tech hiring shift we covered yesterday—where systems design is overtaking traditional resumes—a new dev.to analysis identifies a specific hiring crisis in the AI agent boom. Companies are mistakenly prioritizing deep LLM knowledge over resilience engineering, leading to the hiring of developers who build impressive demos but whose agents fail in production edge cases.
Why it matters
This is a critical, counterintuitive insight for any founder building an AI agent team. The analysis argues that the most important skill is not prompt engineering, but the ability to architect resilient systems that can operate reliably in the face of non-determinism. Founders need to shift their hiring criteria to vet for experience in distributed systems, fault tolerance, and designing for failure. The article provides specific, actionable interview questions to help identify engineers with the right mindset.
The author contends that the 'move fast and break things' ethos is particularly dangerous in agentic AI. A successful agent is not one that works 99% of the time, but one that fails gracefully and recoverably the 1% of the time it encounters an unexpected state. This requires a different engineering discipline than traditional software development.
The Commodity Futures Trading Commission (CFTC) has reportedly broadened its investigation into Polymarket, moving beyond its initial focus on unregistered derivatives. According to reports on Wednesday, the probe is now targeting the exact practices exposed in the WSJ investigation we tracked last month—including the platform's alleged use of undisclosed paid influencers and fake betting videos to attract users.
Why it matters
This represents a significant escalation in the regulatory scrutiny of prediction markets. The focus is no longer just on whether these platforms are legal, but on how they operate and market themselves. The investigation into influencer marketing and deceptive practices signals that regulators are concerned about consumer harm and market integrity. For the industry, this could force a rapid maturation of compliance and marketing standards, pushing platforms to adopt the stricter controls typical of regulated financial services.
The investigation was spurred in part by a letter from senators demanding scrutiny of the platform's practices. Some legal analysts believe this signals a regulatory turf war, where the CFTC is asserting its authority over a new class of products that could also be interpreted as gambling, which is typically regulated at the state level.
The European regulatory net around prediction markets is tightening in line with the ESMA warnings we tracked last month. On Thursday, the Dutch gambling authority (KSA) rejected Polymarket's appeal against sanctions for offering illegal gambling services, maintaining that event contracts constitute gambling under Dutch law regardless of their blockchain-based nature.
Why it matters
This ruling is a significant setback for prediction markets in Europe and highlights a growing regulatory divergence from the US. It establishes a strong precedent that using decentralized technology does not grant an exemption from national gambling laws. Coming on the heels of a pan-EU regulatory crackdown, this decision reinforces the hostile environment for retail-facing prediction markets in the region, potentially forcing platforms to either withdraw or fundamentally restructure their offerings to comply with disparate and strict national regulations.
Legal experts note that this decision could embolden other national regulators within the EU to take similar action, creating a complex and fragmented compliance landscape for prediction market operators. This contrasts sharply with the ongoing federal vs. state battles in the US, where the debate centers on classification as a financial product, not an outright gambling ban.
The newly released Q2 2026 PitchBook-NVCA Venture Monitor formalizes the extreme 'barbell' market data we've been tracking for weeks. While confirming that nearly half of all capital went to just three AI mega-firms, the report adds a concerning new data point: first-time fund formation has hit a decade low as Limited Partners consolidate their capital with established managers.
Why it matters
This report quantifies the 'barbell' market structure we've been tracking. For founders outside of late-stage AI, this is the hard data showing why fundraising is so difficult. Capital is not scarce, it's just hyper-concentrated at the top. This dynamic dramatically shapes what gets built, creating a boom for a few well-funded AI players while starving the broader early-stage ecosystem. It also puts immense pressure on emerging VCs, threatening the diversity of the capital allocation landscape.
The report notes a 'flight to quality' among LPs, who are doubling down on established VC firms with proven track records in the face of a liquidity squeeze. Analysts warn this feedback loop—where large funds get larger and raise the majority of capital—could stifle innovation by making it harder for contrarian or niche-focused early-stage funds to emerge.
The Advanced Research Projects Agency for Health (ARPA-H) has awarded $22 million to Linnaeus Therapeutics to develop LNS8801, an oral drug designed to proactively maintain physical and cognitive abilities during aging. The funding, announced Thursday, supports a shift from treating age-related diseases to preserving 'Intrinsic Capacity' and healthspan before decline begins. The drug, which activates the GPER pathway, was originally developed for oncology but showed promise in improving cardiometabolic health.
Why it matters
This is a major federal investment in a paradigm-shifting approach to longevity: pre-emptive intervention rather than reactive treatment. The funding from ARPA-H, a high-risk, high-reward government agency, signals a growing policy interest in extending healthspan, not just lifespan. For the DeSci and longevity space, this provides significant validation and a potential new funding pathway for ambitious projects that aim to preserve biological function rather than just cure disease.
Researchers in the field see this as a pivotal moment, moving geroscience from the fringes to a central focus of government-backed health initiatives. The drug's mechanism, inspired by observations in women's health outcomes, also highlights the potential for novel insights by re-examining established biological pathways through the lens of aging.
Ethereum co-founder Vitalik Buterin has called for a strategic reassessment of decentralized governance tools, acknowledging declining enthusiasm for DAOs. In a post on Thursday, he argued that in the current 'chaotic' political climate, the focus should shift from rigid, binding on-chain governance to more flexible 'consensus-finding tools.' He proposed using technologies like zero-knowledge proofs to build 'sanctuary tools' that protect collective voice and coordination without forcing premature or divisive formal votes.
Why it matters
This is a significant pivot from one of the space's core thinkers, moving away from the techno-solutionist ideal of fully automated governance. For anyone experimenting with network states or intentional communities, Buterin's critique is a pragmatic reality check. He is advocating for tools that support community texture and consensus-building—the messy human part of governance—rather than just a voting mechanism. This aligns with the observed failure modes of many DAOs and suggests a more mature path forward for decentralized coordination.
Buterin's argument is that in an era of heightened political polarization, forcing binding votes can fracture communities. Instead, he suggests tools that allow for 'soft' consensus, signaling, and coordination can be more resilient and effective. This is seen by some as a retreat from the original vision of DAOs, but by others as a necessary evolution toward more practical and human-centric systems.
A Dedicated Identity Layer for AI Agents Is Rapidly Taking Shape A wave of new products and partnerships from companies like Proof, Enigma, Descope, and Vercel are focused on giving AI agents their own first-class, verifiable identities. This marks a structural shift away from the insecure practice of agents borrowing human credentials, establishing 'Know Your Agent' as a foundational security principle.
Go-to-Market Strategy Shifts to 'Generative Engine Optimization' With B2B buyers increasingly starting their research in chatbots, the new GTM imperative is to be recommended by AI. Multiple analyses today detail the playbook for 'Generative Engine Optimization' (GEO), which prioritizes building ecosystem-wide credibility and structured data that AI models can cite, moving beyond traditional SEO.
VC Liquidity Crunch Concentrates Capital with Established Funds The latest Q2 venture data confirms the 'barbell' market structure we've been tracking. Record headline numbers mask extreme concentration, with a handful of AI mega-deals capturing the vast majority of capital. A persistent lack of distributions to LPs is further squeezing emerging managers and driving capital toward established incumbents.
Ethereum's Institutional Rails Are Being Built in Parallel While Ethereum core devs map a multi-year 'Lean' protocol rebuild, institutional adoption is accelerating on separate tracks. Major players like OCBC, Lido, and Dinari are launching tokenized assets and services on public blockchains, signaling that institutions are not waiting for the final protocol state to integrate.
Prediction Markets Face a Multi-Front Regulatory and Distribution Squeeze The regulatory environment for prediction markets is tightening globally. The CFTC is broadening its investigation into Polymarket's marketing practices, Dutch regulators have upheld a gambling sanction, and Google is now banning Chrome extensions that facilitate real-money trading, creating significant distribution hurdles.
What to Expect
2026-07-10—Interactive documentary performance 'DETACHED OPEN AIR: Transit Zone Sofia' explores migration and belonging at Toplocentrala in Sofia.
2026-08-01—Google's ban on Chrome Store extensions that facilitate real-money transactions on prediction markets takes effect.
2026-09-14—The 7th Longevity Investors Conference begins in Gstaad, Switzerland, focusing on bridging science and capital for longevity research.
2026-10-13—TechCrunch Disrupt 2026 begins, featuring founder-focused sessions on hiring, product-market fit, and scaling.
How We Built This Briefing
Every story, researched.
Every story verified across multiple sources before publication.
🔍
Scanned
Across multiple search engines and news databases
502
📖
Read in full
Every article opened, read, and evaluated
205
⭐
Published today
Ranked by importance and verified across sources
20
— The Distribution Desk
🎙 Listen as a podcast
Subscribe in your favorite podcast app to get each new briefing delivered automatically as audio.
Apple Podcasts
Library tab → ••• menu → Follow a Show by URL → paste