Adversarial interoperability: how startups can compete with tech giants

Executive overview

Dominant platforms use legal tools — software patents, the DMCA, the Computer Fraud and Abuse Act, tortious interference claims — to make compatible products so legally risky that nobody even tries to build them. This wasn't always true: early tech companies routinely reverse-engineered competitors' products to win customers, and the law let them.

The window is narrowing but not closed. A recent Ninth Circuit ruling in hiQ v. LinkedIn found that scraping public data cannot be blocked under the CFAA, offering a legal foothold for founders willing to use it.

Every dominant tech company got there through adversarial interoperability — and now uses the law to prevent others from doing the same.

What adversarial interoperability is and why it matters

• Adversarial interoperability: plugging into a system without permission, to make a competitor's customer your customer.
• Differs from cooperative interoperability (agreed standards) and indifferent interoperability (nobody cares).
• Samba: an Australian grad student reverse-engineered Microsoft's SMB protocol and released it as open source — now bundled in every Linux distro and macOS.
• Apple reverse-engineered Microsoft Office file formats to build iWork, making Macs first-class citizens in corporate environments.
• Both succeeded because the legal thicket didn't yet exist.

The legal thicket that now blocks entry

• Software patents, DMCA circumvention claims, terms-of-service-as-felony (CFAA), and tortious interference have grown up since the 1990s.
• Blizzard and Facebook established the precedent that violating terms of service can be prosecuted as computer fraud.
• Power Ventures scraped Facebook, lost in court on CFAA grounds — timing mattered.
• hiQ scraped LinkedIn; won in the Ninth Circuit; the judge ruled LinkedIn cannot use technical countermeasures to block scraping.
• The jurisprudence is thin and circuit-specific — risk is real but not uniform.

How founders can act now

• The hiQ decision is the key document to show investors when pitching a scraping-based business.
• Mint's model: obtain user credentials voluntarily, log in on their behalf, aggregate data from sites with no API — built a company sold for hundreds of millions.
• A "bot that logs into Facebook as you, fetches your messages, and delivers them in a privacy-preserving context" is a concrete product template.
• Even capturing 2% of a platform's user base is a large business.
• Building such a product creates a legal and normative constituency for changing the law — companies that benefit from reform will fund its advocacy.
• Founders who arrive early may be forced to sell early (Aaron Patzer/Mint), but they still prove the market.

The theory of change: norms, law, and market pressure

• Lawrence Lessig's four regulatory forces: code, markets, law, norms — all four shape what's possible.
• A product users love — even if legally marginal — builds normative pressure for reform.
• Platforms benefit when scraping is banned; VCs invested in those platforms may refuse to fund competitors — but not all VCs will.
• Regulatory capture is real: when an industry concentrates, lobbying positions converge without explicit conspiracy.
• The Bell system analogy: AT&T used state-monopoly arguments to block modem and internet development for decades; its breakup in 1982 enabled the next generation of tech.
• Facebook is positioning itself as a state monopoly — arguing that only a company at Facebook's scale can moderate Facebook's harms.

Copyright: the entertainment framework applied to everything

• Copyright was designed to regulate entities in the supply chain of the entertainment industry — not fans, students, or ordinary internet users.
• The internet only works by making copies; streaming is downloading without a save button.
• Every internet action involves copying; this doesn't make ordinary users part of the entertainment industry.
• Proposed reform: enumerate a floor of always-fair uses (specific durations, line counts), maintain a principled fair-use ceiling for judges to apply.
• The Wind Done Gone precedent: the Supreme Court upheld an unauthorized sequel to Gone with the Wind as fair use, even though it was commercial.
• Deep fakes will require the same flexible principles framework as music sampling — rules written now will need revision as technology evolves.

Market concentration and the epistemological crisis

• When an industry concentrates to a handful of players, collective-action problems become easy to solve for incumbents and impossible for challengers.
• No-poaching agreements among major California tech companies (secret, eventually exposed) illustrate how convergence happens without explicit conspiracy.
• Regulatory capture corrupts truth-seeking processes — Dow Chemical's argument that West Virginians are too obese to be harmed by chemical waste in drinking water is an extreme example of bought science.
• Rising conspiratorial thinking is a rational response to genuine conspiracies and corrupted institutions, not a failure of individual reasoning.
• Cambridge Analytica's claimed "mind control" was phrenology 2.0 — what the platforms actually deliver is targeted finding of people who already hold particular views.

Where to bet as a founder

• The puck is moving toward: platform breakup, adversarial interoperability becoming legally normalised, demand for resilience infrastructure as climate and institutional crises compound.
• Platforms at Facebook's scale cannot serve edge cases (sensitive medical groups, harassment victims, minority communities) without harm — those edge cases are product opportunities.
• Science fiction is a menu of futures, not a prediction engine — use it for idea generation.
• Peak indifference model: problems that build debt over time (climate, surveillance, monopoly) reach a tipping point where the constituency for change grows without external prompting. Founders who are in place when that shift arrives can ride the wave.
• Building for graceful recovery and resilience — not luxury bunkers, but systems that fail well — is both ethically right and a large near-term market.