The AI Frontiers article (reproduced under) builds on a earlier Asimov Addendum article written by Tim O’Reilly, entitled: “Disclosures. I don’t suppose that phrase means what you suppose it means.” I (Ilan) suppose it’s essential to first very briefly undergo components of Tim’s unique piece to assist recap why we—at the AI Disclosures Challenge—care about protocols in AI markets:
On the AI Disclosures Challenge, we’re more and more coming to see disclosures by the lens of networking protocols and requirements. Each networking protocol may also be regarded as a system of disclosures. However these disclosures are way over only a warning label, or a mandated set of studies. They’re a type of structured communication that allows unbiased, decentralized motion.
Tim then notes why this issues for AI’s “market construction”:
The race for first mover benefit by the big centralized AI suppliers like OpenAI and their enterprise mannequin of offering AI entry by metered API subscriptions suggests a hub and spoke railroad design, whereas a world of open weight AI fashions linked by new modes of standardized communication may look extra like a highway system, or at present’s World Vast Net.
…
If we would like a world the place everybody, not simply AI mannequin builders and people constructing on prime of their centralized networks, is ready to innovate and to supply their work to others with out paying a tax to entry centralized networks, we’d like a system of disclosures that allows interoperability and discovery.
On this method, protocols, as a kind of disclosure, can architect more healthy AI markets, not after issues are already too far gone, however by it working as foundational “guidelines of the highway… that allow interoperability”:
Briefly, we have to cease pondering of disclosures as some sort of mandated transparency that acts as an inhibition to innovation. As a substitute, we must always perceive them as an enabler. The extra management rests with methods whose possession is proscribed, and whose conduct is self and opaque, the extra permission is required to innovate. The extra now we have constructed “the rule of regulation” (i.e. requirements) into our methods, the extra distributed innovation can flourish.
Now, on to the primary course!
Protocols and Energy
As AI fashions develop into commodities, tech giants are racing to lock in customers by hoarding their information. Open protocols, backed by open APIs, can facilitate broad information sharing and more healthy AI markets.
Initially printed in AI Frontiers: https://ai-frontiers.org/articles/open-protocols-prevent-ai-monopolies
Can we head off AI monopolies earlier than they harden? As AI fashions develop into commoditized, incumbent Large Tech platforms are racing to rebuild their moats on the utility layer, round context: the sticky user- and project-level information that makes AI functions genuinely helpful. With the suitable context-aware AI functions, every further user-chatbot dialog, file add, or coding interplay improves outcomes; higher outcomes appeal to extra customers; and extra customers imply extra information. This context flywheel—a wealthy, structured user- and project-data layer—can drive up switching prices, making a lock-in impact when gathered information is trapped inside the platform.
Protocols stop lock-in. We argue that open protocols—exemplified by Anthropic’s Mannequin Context Protocol (MCP)—function a strong rulebook, serving to to maintain API-exposed context fluid and to forestall Large Tech from utilizing information lock-in to increase their monopoly energy. Nonetheless, as an API wrapper, MCP can entry solely what a specific service (comparable to GitHub or Slack) occurs to show by its API.
To totally allow open, wholesome, and aggressive AI markets, we’d like complementary measures that guarantee protocols can entry the total spectrum of consumer context, together with by:
- Assured entry, for licensed builders, to user-owned information, by open APIs at main platforms.
- Transportable reminiscence that separates a consumer’s agentic reminiscence from particular functions.
- Knowledge utilization guardrails governing how AI companies can leverage consumer information.
Drawing on the instance of open-banking rules, we present that safety and information requirements are required for any of those proposals to be realized.
Architecting an open, interoperable AI stack by the protocol layer is about supporting broad worth creation fairly than worth seize by a couple of corporations. Coverage efforts comparable to the EU’s Common-Objective AI Code of Follow do matter; however, in the end, it’s software program structure that almost all instantly and decisively shapes market outcomes.
Protocols—the shared requirements that allow totally different methods talk with each other—perform as a deeper de facto regulation, enabling unbiased, decentralized, and safe motion in digital markets.
From Commoditized Fashions to Context-Wealthy Functions
From fashions to companies. In a fevered race to blitzscale its solution to AI dominance, OpenAI took an early lead. ChatGPT turned the fastest-growing utility in historical past, and it was straightforward to imagine that the following step was to show it right into a platform. OpenAI tried to develop into a developer platform, first with plugins after which with its GPT Retailer.
But it surely hasn’t all gone in response to plan. OpenAI’s fashions don’t appear so particular anymore. Open-source fashions like Kimi K2 (by Moonshot AI) have aggressive capabilities and are free to make use of. Sensing the turning tide, application-specific corporations like Perplexity struck gold by taking off-the-shelf fashions from a number of suppliers, scaffolding them for particular makes use of, and charging for premium entry whereas avoiding vendor lock-in. Cursor, an AI‑first code editor, went from $0 to over $100 million ARR in 18 months, proof that context‑pushed retrieval-augmented era (RAG), with a local AI design, can beat incumbents sitting on extra consumer information. Entrance-end customers can now simply select their most popular mannequin inside these functions. And, utilizing platforms like OpenRouter, builders may even swap fashions dynamically in response to pricing or options.
Context rising. As basis fashions commoditize, competitors is shifting up the stack, to the applying layer, the place proprietary consumer and mission information—referred to as context—is the key sauce. Tech giants are racing to surround and personal this context solely: dialog histories, reminiscence shops, workspaces, codebases, paperwork, and the rest that helps their brokers predict and help higher. OpenAI, Google, and different mannequin distributors lean on chatbot interplay logs as sources of persistent reminiscence, whereas utility specialists like Anysphere (which makes Cursor) and Perplexity equally harness mission and consumer information to spice up their fashions’ usefulness.
This forces an important resolution in the marketplace: will AI functions develop based mostly on closed requirements that allow a couple of gatekeepers dictate phrases and extract outsized rents, or on open requirements that preserve context moveable and structure permissionless?
The early open internet. The stakes are excessive. Born on open protocols, the net advanced into ecosystems of functions dominated by Amazon, Google, and Meta. At first, they beat rivals just by working higher. Google was one of the best at matching searchers with data and advertisements; Amazon surfaced one of the best merchandise at low costs; and Fb matched its customers with a singular feed crafted solely from content material shared by their pals and folks they selected to observe.
From innovation to extraction. However success conferred sturdy energy that was abused. As development slowed, the profitable corporations shifted from creating worth to extracting it. In our previous work, we described this course of utilizing the language of financial rents: winners first achieve “Schumpeterian rents” for innovation, however, as soon as markets mature, these flip into extractive rents geared toward preserving dominance and squeezing customers and builders. Cory Doctorow frames this course of vividly as “enshittification.” AI’s enshittification may contain weaker security guardrails, greater costs, much less consumer privateness, and lower-quality data or agentic help. Briefly, when business incentives go unchecked, fashions get tuned to serve suppliers’ pursuits over these of customers.
Makes an attempt by OpenAI to construct a platform by locking in builders and customers resemble Fb’s failed try and construct a platform. However, as Invoice Gates is stated to have commented: “This isn’t a platform. A platform is when the financial worth of everyone that makes use of it, exceeds the worth of the corporate that creates it. Then it’s a platform.” That sort of platform is sort of all the time enabled by open requirements. Against this, when an organization blocks others from linking appropriate merchandise to its ecosystem, it incentivizes clients to make use of a number of companies directly (`multi-homing’), and invitations additional scrutiny from regulators.
The promise of protocols. Anthropic has taken a distinct route, creating MCP as an open protocol, a shared algorithm that anybody can use free of charge. MCP standardizes how AI functions request data and actions from exterior companies, thereby facilitating equitable developer entry to exterior instruments and information context. That is how networked markets develop: by enabling an structure of participation by which each and every new entrant makes the market extra precious for everybody else.
MCP’s take-up has been explosive. As we speak there are nicely over 5,000 MCP servers that can connect with the tons of of AI apps which have built-in MCP. Confronted with fast adoption by third-party builders, AI mannequin builders like OpenAI and Google have introduced that they too will assist MCP. However these identical incumbents are already pushing again.
How Person Context Is Powering a New Period of Tech Monopolies—and Competitors
Context creates worth. AI methods thrive on context: the consumer information that lets an AI system tailor its conduct to customers, their requests, and the duties at hand. When correctly mined, this consumer information permits for personalised and environment friendly predictions. Consider a context-free, factory-settings AI mannequin as a borrowed cellphone: the {hardware} is highly effective, however, with out your contacts, messages, location, and logins, it may well’t actually assist you to.
Context has many layers: throughout time, as a residing “state,” such that every consumer immediate builds on what got here earlier than; and throughout individuals, as a multi-user setting (say, in a Slack thread or collaborative doc). We emphasize two layers: micro-context captures whom the system helps proper now (related to their preferences, language, and present question). Then again, macro-context covers the duty surroundings, because the exterior body that shapes what a smart reply seems like. This consists of mission recordsdata and dwell information feeds.
Large AI corporations are utilizing context to develop their moats and lock in customers by no less than two approaches. The primary is thru product bundling. Examples embrace OpenAI’s push into search, analysis, and coding (together with by acquisitions); Google’s threading Gemini into Workspace; Microsoft’s embedding Copilot throughout its 365 productiveness apps. Bundling aggregates the information floor and raises switching prices.
The second is thru constructing context as a central product characteristic. OpenAI now gives persistent reminiscence that shops private particulars (e.g., “has a toddler” or “identified with ADHD”) to form future replies. Meta has introduced it’s going to accumulate cross-site consumer information to personalize its AI assistants. Google now remembers your writing type, so it may well tune its AI-generated Gmail replies. By binding the app and its context to the mannequin, corporations lock in customers and starve rivals. Such bundling is fertile floor for enshittification.
Importantly, this course of depends on Large AI corporations’ gathering express consumer indicators—their prompts, docs, API calls—and distilling them into an inferred, implicit preferences profile that lets their mannequin ship extra related, environment friendly predictions inside every consumer’s distinctive workspace.
Can Protocols Create a Degree Taking part in Area?
The MCP pipeline. Anthropic’s MCP standardizes how AI functions request instruments, information, and actions from exterior companies by a common adapter. As a substitute of customized integrations for every pairing (Cursor → GitHub; Claude → Google Drive), any AI app (each an MCP shopper) can use any MCP-compatible service (or MCP server), making fashions extra interchangeable. MCP additionally creates an agentic interface that enables an AI agent to resolve what to do, based mostly on the language of duties, not endpoints. This reduces the MxN integration tax, permits small corporations to hire fairly than construct tooling, and weakens vertical exclusives.
As a result of MCP is client-agnostic, any AI app can use any exterior service, which in flip makes switching between fashions far simpler — both by switching between mannequin service suppliers that assist MCP, or by constructing an unbiased MCP shopper and utilizing any mannequin service. When an AI app’s context is moveable, fashions develop into extra interchangeable.
MCP is the last word unbundler of context: any appropriate AI app can attain any service that exposes an MCP server, permitting an enriched immediate to then be despatched to the mannequin. However companies should nonetheless decide in, by making their content material accessible by APIs.
This shifts the aggressive gravity “up the stack,” away from the mannequin builders and to the applying that develops the profitable context flywheel. App-level information portability and governance—together with pricing, permissioning, and any preferential entry into Large Tech–managed information sources—then turns into the brand new battleground.
Though MCP reduces integration friction, interoperability alone doesn’t guarantee market competitors. We’ve seen this earlier than: open protocols like HTTP (for internet searching) and SMTP (for e-mail) enabled permissionless entry of latest functions, but markets nonetheless tipped. Google is now the dominant e-mail and browser supplier due to its superior merchandise and cross-app integrations.
MCP’s Impression on the AI Market So Far
Incumbents have rushed to insert AI into each legacy product: the quickest go-to-market technique with the shallowest integration. Meta surfaces an assistant in almost each app. This has solely made constructing cleaner, MCP-enabled functions much more engaging. AI-native instruments like Perplexity supply additional encouragement to builders, exhibiting that customers will decide a custom-made expertise over a retrofitted one (just like the AI-layered Google Search).
Unsurprisingly, the variety of new MCP servers has rocketed, as we famous earlier. Nonetheless, such integrations might also be boosting utilization of incumbent mannequin builders’ chatbots as they achieve entry to extra instruments. MCP’s affect has been impeded by its weak safety. MCP servers’ exterior authentication and authorization stay a cussed MxN integration drawback. Furthermore, for repeated manufacturing workflows, code-based frameworks could also be extra environment friendly than an inference‑solely workflow.
Lastly, there are early indicators that AI mannequin builders could resist interoperability extra broadly, regardless of the elevated utilization it generates for them, if it finally ends up reinforcing the context moats for utility builders. Anthropic briefly minimize off the coding utility Windsurf’s direct (first-party) entry to its high-performing Claude fashions. Windsurf was rising too well-liked and was set to be acquired by OpenAI, a direct competitor to Anthropic.
MCP Versus Walled Gardens: The API Gatekeeping Downside
APIs are the gateway by which an MCP shopper—the AI functions—can entry third-party information and instruments, thereby breaking down a platform’s “walled backyard” of proprietary companies and datasets. However MCP can liberate context solely when a third-party service gives a sufficiently wealthy API (and retains it open). As a result of platform homeowners management these APIs, they’ve an incentive to constrain what MCP can contact, to guard their aggressive edge. This manifests in two methods:
- Entry danger. Companies can merely shut off API entry completely, or they will drastically degrade entry. Latest API paywalls and shutdowns at Reddit, Twitter, and Meta present how entry can vanish in a single day. Enterprise companies like Salesforce (which owns Slack), Atlassian, and Notion are now limiting API entry by Glean (a context platform) at the same time as they launch competing merchandise. In the meantime, Slack’s new API modifications (supposedly to restrict how LLMs are in a position to entry the app) will hurt builders generally.
- Context-depth danger (the “personalization hole”). Platform APIs expose posts and recordsdata however not often the behavioral profiles that energy their very own personalization, leaving newcomers with a chilly‑begin handicap. Meta, for instance, personalizes its personal chatbot with Fb and Instagram historical past, nevertheless it gives third events neither its Graph API to fetch that full profile nor entry to detailed points of customers’ express and implicit (inferred) profiles. Equally, OpenAI’s “reminiscence” characteristic is confined to ChatGPT. OpenAI doesn’t enable builders to entry a consumer’s “recollections” by way of an API, even with the consumer’s prior consent.
To Save AI from Enshittification, Assist Protocol-Degree Interventions
Enhancing protocols for the AI age. To interrupt API gatekeeping in AI markets, we’d like an structure that helps user-sanctioned information portability so as to improve third-party developer entry. Right here, portability means finish customers’ capacity to learn and switch their information throughout platforms—or to permit different builders to take action on their behalf. When portability is common, builders can entry the identical context (by MCP or any API) with out negotiating bespoke offers. To operationalize this method for AI markets, we suggest:
- Open API entry for main platforms. If the information comes from the consumer, the consumer—and any developer the consumer authorizes—ought to have the ability to take it elsewhere. We suggest requiring that, with consumer consent, main platforms expose this user-owned contextual information by APIs to accredited builders at zero value. We suggest beginning with the platforms that management essentially the most consumer context: “gatekeepers” designated by EU standards, plus main AI mannequin suppliers.
Such an method may draw inspiration from the EU’s open-banking regulation (particularly, its Second Cost Companies Directive, or PSD2), which holds that banks should present licensed fintechs with free, real-time entry to core account information and fee capabilities. Licensed builders should first receive a license by exhibiting correct safety and information requirements. In contrast to banking’s standardized information, although, AI context spans code repositories, conversations, behavioral patterns, and preferences. Within the case of AI, markets and regulators would wish to provide you with a manner of defining “core consumer context” for these varied information sorts and platforms.
- Reminiscence as a conveyable service. Customers’ AI “reminiscence” must be accessible throughout platforms by way of APIs, with market-driven safety requirements embedded within the technical structure. Such MCP servers already exist, even when AI functions don’t assist it.
The problem is much less technical than socio-economic. Reminiscence is deeply private and requires safe data-handling, but AI markets at the moment lack requirements and accreditation in these areas.
A market-driven method can be to embed these safety requirements into technical structure, as is completed with the FDX API commonplace for US open banking. Such embedding permits for safe and standardized sharing of economic information between banks and third-party builders. Safety necessities like end-to-end encryption, OAuth-controlled entry to client-side keys, and granular topic-by-topic permissions are at the moment past MCP’s scope. However FDX’s safe and common API reveals what is feasible.
- Secure personalization, with out information exploitation. Open APIs rely on customers’ trusting builders to deal with shared context responsibly. Trade-specific information utilization guidelines would additionally weaken incumbents’ benefits whereas creating safer applied sciences. Such utilization guidelines may begin with:
- Knowledge firewalls. We suggest defending intimate consumer conversations from business concentrating on. An AI utility leveraging a identified consumer choice like “is vegetarian” for restaurant suggestions is useful; however exploiting therapy-like conversations for manipulative promoting have to be prevented.
- Erasure rights. Customers ought to have the ability to evaluate, edit, or delete their choice profiles and recollections at any time. ChatGPT already largely gives this.
- Privateness defaults. For delicate queries, we suggest that AI companies default to a non-public mode, with out long-term reminiscence enabled or advert concentrating on, except customers explicitly decide in to those settings for such queries.
In the end, management over consumer context—not uncooked mannequin energy—will resolve who wins the AI business race. Open protocols can preserve context fluid between opponents, however they’re solely as efficient as the information (and instruments) that they will securely entry. The selection is ours: design aggressive AI markets round open ideas, or settle for a brand new era of platform monopolies.
Because of Alex Komoroske, Chris Riley, David Soria Parra, Guangya Liu, Benjamin Mathes, and Andrew Trask for studying and/or commenting on this text. Any errors are ours.