A statewide goal for New York
Make the first stop the last one anyone needs.
Mission Zero is a plan to drive every kind of sexual assault in New York toward zero, with a special focus on date and acquaintance rape. It isn't only about college. Risk across the ape-and-rape spectrum follows everyone who dates, from senior prom to a first marriage before or after 30, and beyond it: working adults, people who never went to college, and New Yorkers dating again at 18 or 81. It treats assault the way we treat fire: with prevention built into rooms, phones, parties and law, tested in a statewide trial and measured honestly.
The problem
The system notices too late
Most sexual assaults are not attacks by strangers. They happen between people who know each other, on dates, after parties, in bedrooms, on and off campus. We respond with rape kits, investigations and hearings. All of it starts after the harm. And reporting is rare enough that one person can harm several others before anyone files a report.
Everyone who dates. Campuses get the headlines, but most people at risk aren't students. Date-rape risk starts for many around senior prom and lasts through every stretch of single life: the twenties after school or instead of it, dating again after a divorce or a death, and new relationships at 50, 60 or 81. A bartender in Buffalo, a nurse on night shifts in the Bronx and a retiree on a dating app in Albany deserve the same protection as a sophomore in a dorm. Whether or not someone can afford college, wants it, or has long since finished it, they deserve a minimal bedroom safety standard.
At the 20% rate at which college women report rape and sexual assault to police, someone must harm about 3 people before any report is more likely than not, and about 10 before one is nearly certain (Bureau of Justice Statistics, 1995–2013). Gray-zone pressure, such as persisting after a no, is reported far less, and no one measures it, so the lower rows below are illustrations.
| Chance each person reports | Harmed before a report is more likely than not | Harmed before a report is 90% likely | Chance of any report after 5 people | After 10 people |
|---|---|---|---|---|
| 32% (non-students, assault) | 2 | 6 | 85% | 98% |
| 20% (students, assault) | 3 | 10 | 67% | 89% |
| 10% (illustrative) | 7 | 22 | 41% | 65% |
| 5% (illustrative gray zone) | 14 | 45 | 23% | 40% |
| 2% (illustrative) | 34 | 114 | 10% | 18% |
Method: the chance of at least one report among n people is 1 − (1 − p)ⁿ, assuming each person reports independently at rate p. A report is also not a consequence: few police reports lead to arrest, reports about the same person often land at different schools or apps and never connect, and a minority of repeat offenders commit most assaults.
The design gap
Bedrooms were never designed for the harm that happens in them
A dorm room, a studio apartment or a hotel room comes with a smoke detector, a marked fire exit and often a sprinkler head and a carbon monoxide alarm. It comes with nothing for sexual assault, even though more than half of sexual assaults happen at or near the victim's home (RAINN). Whoever designed these rooms was, on this point, an idiot: they engineered them for fire and left out the harm so many people who date fear most. The people who design a room's furniture, surfaces and standard devices bear responsibility for what those things could have prevented and didn't. That is a design failure, and design failures can be fixed.
The technology isn't new. Building intercoms were patented in 1894, and two-way intercoms were common in apartment buildings by the 1950s. Hospitals have wired call buttons at every bed, and affordable hardwired panic buttons, which could be framed for the gray zone or the red zone, have been ordinary parts of security systems for decades. That dorms, apartments and other buildings weren't built around this kind of safety was a choice: a choice to neglect one of the greatest safety risks people face from their teens through every later season of dating. Someday we need a historical examination of exactly why bedrooms were never engineered to protect women and men from assaults that go uninterrupted because a person freezes, goes still or fawns.
I've been Semmelweissing this since 2015. Like Ignaz Semmelweis, who showed in the 1840s that handwashing saved mothers' lives and was ignored for years, I've argued on and off since 2015 for building sexual assault safety into bedrooms. Despite my best efforts, the message hasn't changed campus housing standards, let alone the standards for apartments, rentals and hotels where most adults sleep. Perhaps after the uproar over the Cornell case, campus housing authorities, landlords, builders and those who influence them will finally build bedrooms with clear, direct safety features for sexual assault prevention, using modern electronics that are privacy-engineered and worthy of people's trust.
Part of the reason may be fear of how it looks. A college that builds a help button into its bedrooms worries that parents will read it as an admission that its dorms are dangerous. I once met a parent who took one campus's gift of a personal emergency bracelet to every incoming student as a sign that the campus was especially dangerous, when it was really a sign that the campus was especially careful.
The answer is for every campus to move at once. When New Jersey required sprinklers in every dorm after the 2000 Seton Hall fire, no campus looked dangerous for having them. If every New York campus adopts the same housing safety standard at the same time, with unobtrusive, privacy-first technology in the same spot in every bedroom, it stops looking like a confession and becomes a baseline, like a seatbelt in every car. Then parents and visiting high school students will ask why a campus doesn't have it yet. The same logic holds for landlords, hotels and home builders: when the standard is statewide, renters and guests will start asking why a bedroom doesn't have it. The proposals for this are P-39, P-41, P-43 and P-77.
A device is a chance to de-vice. Every device in a bedroom, from a speaker to a lock to a headboard, is a chance to reduce vice: coercion, pressure and harm. The goal is to diminish vice, not sex. If anything, safe sexuality should grow when people feel protected enough to express themselves and explore within safe bounds, under ground rules both partners understand, in caring and loving relationships.
For every couple. Mission Zero is pro-LGBTQ. Every tool and proposal works the same for partners of any gender or orientation, never requires anyone to reveal their orientation to anyone else, and treats every consenting relationship equally. CDC surveys find that lesbian, gay and bisexual people face sexual violence at higher rates than their straight peers, so they stand to gain the most from rooms built to protect everyone.
The Semmelspan
How long we have known, and waited
The Semmelspan is the time between when a problem and its fixes are made public and when the system finally acts. These clocks are still running. Each marks a moment when date rape and the means to interrupt it were put plainly in front of the country, and bedrooms still weren't built to stop it.
1988
I Never Called It Rape is published
Robin Warshaw’s Ms. report brings date and acquaintance rape on campus into public view.
Semmelspan: about 38 years
September 19, 2014
“It’s On Us” launches at the White House
President Obama and Vice President Biden launch the national campaign against campus sexual assault.
Semmelspan: 12 years, 44 days
October 1, 2014
I describe this system to the Small Business Development Center
I presented the proposed technical system to the Small Business Development Center, under the auspices of the U.S. Small Business Administration, at the University Heights Library Business Development Center. Watch the recording.
Semmelspan: 12 years, 32 days
September 21, 2015
The first major AAU Campus Climate Survey is released
The Association of American Universities publishes results from students at 27 universities.
Semmelspan: 11 years, 42 days
July 7, 2016
My application is published
U.S. Pub. No. 2016/0196735, filed under the name Adam Clayman, becomes public.
Semmelspan: 10 years, 118 days
A different night
How a night could go differently
Here is how a date or party night could unfold with systems that can be built and adopted today. Campuses are the easiest place to test them, but the same layers work for anyone dating, in apartments, hotels and homes anywhere in New York. Each step is a layer; none depends on the person at risk recognizing danger in time.
At the door
A registered party runs breath tests at entry and marks served guests, as bars already do with wristbands. Free drink-test strips sit at every bar. Above a set breath-test line no one can consent, and in a group encounter that bars everyone in the room.
Arriving somewhere unfamiliar
If someone arrives late at night at a residence their phone has never seen, after drinking or with a new date, the phone raises one quiet, private check-in: fine, "check on me," or help. Nothing leaves the phone unless they ask.
Before anything starts
If either partner may have had more than one drink for a woman, or two for a man, each has a duty to know the BAC picture: the instant reading and what it means, whether each can consent right now, and if not, when they will be able to. A breath sensor beside the headboard button shows all three, because blood alcohol keeps rising after the last drink. Without a test, the answer is to wait.
At the first "stop"
A press of the headboard button, or a universal safeword spoken aloud and trained as reflexively as stop, drop and roll, sets three things off within seconds. The phone or speaker, listening on the device for that one word and nothing else, stops the music and plays a calm check-in. A silent alert with the room number goes to a trained, sober responder nearby: a peer on campus, a vetted neighbor or a trusted contact anywhere else. If there's no answer within a minute, campus safety or 911 gets the room-level location too.
Minutes later
A responder knocks: "Welfare check." A knock can end an incident without anyone being arrested, and it arrives while harm can still be stopped. If the person at risk wants to leave, a refuge for the night is assigned in minutes: a lockable safe-sleep room or a vetted host, opened with a time-limited key, never revealed to the person left behind.
The morning after
Every attendee gets one optional, anonymous "how was last night?" message. A survivor can lock a time-stamped account in a private evidence vault, and file a sealed name that unlocks only if someone else names the same person.
The proposals
115 proposals across 12 themes
These are options, not a package to adopt whole. Each is a short summary, not yet a bill. The tag on each marks where Lifesaver Labs has a financial stake. Cross-references like P-64 are links.
The pilot
The first statewide Mission Zero trial
New York would be the first state to run Mission Zero statewide: a three-year, campus-randomized trial across participating colleges and training programs, testing which combinations of policy, environment and technology cut sexual assault the most per dollar. Each arm adds a layer to the one before, so the state sees what each added step buys. Its size follows what New York chooses to include, and a focused start can be inexpensive.
Campuses are where the trial starts, not where the problem ends. Schools can randomize, survey and measure in ways the wider public can't, so they make the cleanest first test. What works there is meant to reach everyone who dates: the same button, safeword and check-ins in rentals, hotels, homes and dating apps, for New Yorkers who never enrolled or graduated decades ago.
| Arm | What campuses get | Main proposals | Added cost |
|---|---|---|---|
| 0 · Comparison | Current practice, plus the shared measurement every arm gets | P-46, P-54, P-50 | Survey and reporting only |
| 1 · Culture and education | Ask-first norms, a first-weeks course, group-chat ethics, resistance and bystander training, a confidential help line | P-19–P-23, P-27, P-44 | Low: staff time and curriculum |
| 2 · Plus safer parties | Sentinel-event triggers, breath testing and served-guest marking, drink-test strips and chapter scorecards | P-32–P-38 | Low: mostly policy |
| 3 · Plus touch-and-voice basics | A headboard button in a standard spot in every student bed, room-level location, phone-SOS training and secure locks | P-39, P-43, P-84 | One-time: about $20 per button, reused with the bed, plus lock retrofits |
| 4 · Full integrated platform | On-device stop-word detection on students' own phones, signals routed to trained peer responders, morning-after check-ins | P-44, P-45, P-76–P-78 | Low per student; ongoing cost is mostly trained responders |
- Randomize whole campuses, balanced by size, residential share and Greek presence. Comparison campuses receive the best arm when the trial ends.
- Measure every arm the same way. The primary outcome is an anonymous survey, identical everywhere, because better reporting systems raise reported numbers.
- Independent and open. An independent evaluator preregisters the trial and publishes all results, including null ones. Lifesaver Labs and other vendors have no role in evaluation.
- Safety stops. Any arm stops early if help calls or harms rise.
Privacy
A slightly different balance, with hard guarantees
This asks people who date to accept a button above every bed and a phone that listens for one word. We accepted smoke detectors in every bedroom because they do one job and report nothing else. Safety tech for the bedroom has to meet that bar, in law and in code.
- Detection happens on the device, including reading the affirmative signals of the consent envelope. Nothing is recorded or sent unless the person triggers it.
- Open standards with several certified vendors, so no single company controls the system.
- Published data flows and verifiable builds, so anyone can check the software does only what it says.
- Pressing a button or using any tool is always the person's choice. Declining carries no penalty and no adverse inference.
- A record of a yes from any tool can never be used as proof of consent.
- No cameras and no brain-signal data in intimate-safety devices.
- Records stay in the user's custody, encrypted to keys only they hold.
- Statistics leave a campus, building or app only as de-identified totals, in groups large enough that no one can be identified.
Privacy and capability
Choose how much the room can understand
Every bedroom gets the same baseline, and each person chooses how much more to turn on. Slide from the least processing to the most. Each step adds what the room can notice, and asks for more trust in local, on-device processing. Nothing at any level is recorded or sent unless someone asks for help.
Gray-zone button
Trust needed: minimal- What it does
- What it processes
- What stays private
- What it can catch
Many designs, all testable. The consent envelope isn't one product. It is a family of configurations that can be designed and tested in the pilot, each set by the people using it:
- Safeword only: listens for nothing but the universal safeword.
- Cue words: also listens for a small, preprogrammed set of words the partners chose, such as go, pause, slow down and stop.
- Deeper reading: a private on-device model interprets how things are going from how partners interact, never from body signals alone.
- Private intimacy coordinator: at the far end, an electronic intimacy coordinator you can talk to beforehand to scope out what you want and don't want, which then supports you to the max during the moment, all on the device.
How the envelope carries consent: the music's loudness follows the quietest partner's affirmations.
Levels 2 to 6 use patent-pending mechanisms, licensed on low-cost, fair terms (see the pledge). Every level keeps the button and the safe phrase, and stop always wins.
Already here
Tools on the market or coming soon
Much of a Mission Zero setup can be assembled today from products other companies already sell. Every setup should offer both a touch path and a voice path: voice works when hands are held or a phone is out of reach, and touch works when speaking is unsafe. Mission Zero's standards are open, so any of these can be certified and used.
No new hardware required. Every piece can be off the shelf: phones, watches, smart speakers, programmable buttons and breath sensors that are already sold today. New hardware engineering is the slowest and hardest part of any safety system, harder than software and platform integration, and Mission Zero doesn't depend on it. What holds this back is focus: a community spotlight, and intense cooperation among industry and institutions, both collaborative and competitive, aimed at the Mission Zero goal. It takes a willingness to take responsibility for the engineering fixes, and to withstand sometimes wild public criticism and early rejection while public understanding matures, until everyone is comfortable with minimally invasive, privacy-centered "bedroom seatbelt" technology for high-voltage moments across the hot-cold empathy gap.
Why platforms wait. The first-mover problem that keeps one campus from adding bedroom safety features also holds back the biggest platforms. A company like Amazon may not want to be the one to spearhead a universal, freely available bedroom safeword on every Alexa device owned by Prime members, for fear of an unpredictable customer backlash or boycott spilling over onto its core shopping business. No company wants to move alone on this, and no individual inventor or small business can make a platform prioritize options to prevent sexual assault in Prime members' bedrooms, students or not, however clearly the need and the capacity are laid out. That changes when Prime members, institutions, the public and the New York State Assembly ask for it together.
| Product | Touch | Voice | What it does | Cost |
|---|---|---|---|---|
| Phone emergency SOS (iPhone, Android) | Repeated side-button presses | “Hey Siri” or “Hey Google, call 911” | Calls 911 and texts emergency contacts with location | Free, already owned |
| Smartwatch SOS | Button hold; fall detection | Assistant voice command | Calls 911 from the wrist when the phone is out of reach | Free, already owned |
| Programmable panic button, such as Flic | Single, double or held press, each mapped to a different action | None | Sends a text, push alert or SOS with location; hubs can give room-level location | About $20 each |
| AI voice button: Flic Cue | Press, double press, hold, swipe or turn, each mapped to a different action | Hold and speak; the microphone is on only while held | Can be set up as an AI-sensitive panic button: a whispered “I want to leave” can start a quiet check-in or a ride, while “get help now” escalates to an alarm. Routes speech to messages, a webhook or an AI agent through a nearby phone, computer or Flic Hub. Processing appears to be cloud-based, so it should not be the only path | Kickstarter from November 2026, shipping February 2027; price not yet listed |
| Safety jewelry: invisaWear | Discreet press on the jewelry | Secret phrase through its app | Alerts emergency contacts; a subscription adds 24/7 ADT agents who can send help | $129, plus an optional $11.99 a month |
| Phone-free wearable: Loris SOS, as a bracelet, scrunchie or keychain | Press both side buttons together; five clicks skips the delay | None | After a cancelable 30-second countdown, alerts 911 with name, location and floor; no passive tracking; works without a phone | University pilots and a consumer waitlist; price not yet listed |
| Smart speaker emergency service: Alexa Emergency Assist | None | “Alexa, call for help” | Connects to a 24/7 agent who can dispatch police, fire or EMS; cloud-based, so it fails offline | $5.99 a month |
| Campus safety apps, such as Rave Guardian and LiveSafe | In-app panic and timer | Varies | School-run alerts, safety timers and anonymous tips | Institutional license |
| Safer Yes (Safeword) and Neighbor 911, by Lifesaver Labs | Touch fallback, panic and app alerts | On-device stop words and affirmation windows | Lapse leads to a check-in, alarm and help without the cloud; alerts route to vetted nearby responders | In development |
A button that understands. Most panic buttons know one thing: that they were pressed. A hold-to-talk AI button like the Flic Cue can also tell a gray-zone moment from an emergency, so a whispered “I’m not sure about this” gets a private check-in instead of a 911 call, and a clear call for help skips straight to one. Because the microphone listens only while the button is held, it adds understanding without an always-on ear. It fits naturally at the center of the headboard, as the touch and voice path in one device.
Prices and availability as last checked; confirm with each company. Listing a product is not an endorsement, and none of these companies is affiliated with Mission Zero NY.
The tools
Built for the moments outside the session
Most safety tools wait for a person to recognize danger and act. These are designed for the moments where that fails: before a decision is made in the heat of the moment, at the point of exit, during a frozen dispute, and in the bedroom itself. Each is patent pending and will be licensed on low-cost, fair terms so that patents never block adoption (see the pledge).
In the bedroom, by choice · Patent pending
An affirmative consent envelope
For partners who opt in, often early in a relationship, music plays only while each person keeps giving a simple sign that things are going well: a tap, a squeeze, or a word they enrolled. If the signs stop, the music fades and a private check-in follows. Consent becomes something the bedroom helps people check on, instead of something assumed to be on autopilot.
That matters because some of the most common reactions to pushed boundaries are quietness and tonic immobility, the involuntary freeze many survivors describe. The wish to stop often forms long before a person says so, and then they may say it softly. A fading song notices the silence. This asks for more trust than the guard state, because devices have to read affirmative, enthusiastic signals, so that processing stays on the device and nothing is recorded or sent. Body signals never count as a yes, either partner can privately drop the envelope as trust grows, and the safe phrase stays armed throughout.
In the bedroom · Patent pending
A guard state armed by an ordinary cue
A song, a routine voice request or a normal device command quietly arms the room's devices, with no visible change. They listen on the device, without a wake word, only for a safe phrase, and keep no other audio. A safe phrase starts a private check-in, then an alarm, then a call for help, each only if there's no answer. False alarms are filtered before anyone calls 911, and any request for emergency help carries a confidence level.
On arrival · Patent pending
A late-night arrival check-in
Apple's and Google's safety check-ins confirm you got home and treat arrival as the end of risk. This one is raised by arrival itself: late at night, at an unfamiliar residence, in a risky context. It's private, dismissible in one gesture, and has a duress answer that looks like an ordinary dismissal.
At the point of exit · Patent pending
A refuge for the night
Someone who wants to leave a partner's or host's room at 2 a.m. often has nowhere safe to go. A discreet request, by button, check-in answer or phrase, gets screened for medical risk first, then matched to a safe-sleep room or a vetted host with no possibility of romantic interest, and opened with a time-limited credential.
After a report · Patent pending
A held-apology and clarification channel
A no-contact order protects the complainant but freezes every misunderstanding in place. Through an ombudsperson, or any office entrusted with the role, a restrained party can submit an apology held for the protected party to open only if they choose, and ask closed, structured questions about the investigator's understanding. Each side can verify what was transmitted, changed or withheld, without relying on the intermediary's word.
During a no-contact order · Patent pending
Distance without surveillance
Phones that hold a shared secret can tell when the two parties are about to cross paths, by Bluetooth or private location matching, and gently suggest a different route, venue or time. Neither learns where the other is. Class and work calendars forecast conflicts early, and both can periodically ask for more distance or less.
Everywhere · Patent pending
One universal safeword
New York adopts one emergency phrase that every voice-capable device honors, whether or not the user set one up, and an open interface so safety apps can stop playback and confirm the device complied. Stop always wins over any other command.
Patents
Patents will not be the barrier
Some mechanisms here are covered by U.S. Patent No. 12,347,301 and by provisional patent applications. A provisional is a one-year placeholder no examiner has reviewed, and many of those claims may never become a full application.
The cautionary tale is SawStop. A table saw that stops its blade on contact with skin went on sale in 2004, after manufacturers turned down its inventor's licensing offers. Manufacturers later cited its patents in fighting a federal safety standard, and more than 20 years on that standard is still not in force, while about 30,000 table-saw injuries a year needed medical treatment, about 4,000 of them amputations. A patent thicket in this field could let hundreds of thousands of preventable assaults happen nationally.
The better model is the three-point seatbelt. Nils Bohlin designed it at Volvo, which patented it in 1959 and then left the design open for every carmaker to use. Some of the claims behind this work, together with a few other inventors' claims in this field, may matter as much for bedroom and intimate safety as that patent did for cars. That is why they must not become a toll booth.
One of those other claims needs naming. U.S. Patent No. 12,512,094, “System and method for consent detection and validation,” was granted to Wassim Chaar on December 30, 2025, on an application filed July 26, 2022, six years after my own application was published. It describes detecting consent or dissent from speech and sensor data during a scheduled in-person meeting.
- Why I believe it may be challengeable. My 2016 publication was listed in its file only after allowance and never applied by the examiner, though it describes the same field: safewords, voice recognition, sensors and summoning help. I’ve prepared a request for ex parte reexamination arguing that its claims are anticipated by or obvious over that publication, alone or combined with other art the examiner didn’t consider.
- Why I believe it doesn’t reach the baseline here. Its claims were allowed for one reason: a confidence score built from both people’s intent, current sensor data and characteristics predicted for both of them. Recognizing one fixed safe phrase computes no such score.
- What that means for the State. An issued patent is presumed valid until the Patent Office or a court decides otherwise. This is my position, not a legal determination, and the State should get its own review before relying on it.
- Anything here that turns out to be my invention will be licensed on low-cost, fair, reasonable and non-discriminatory (FRAND) terms, to industry, institutions and other innovators alike.
- Where price would otherwise block adoption, by industry or by a trial community such as Cornell, the SUNY system or the State of New York, I'll go lower: roughly the cost of the patent filings plus two years of forgone salary, or much less.
- No proposal here will go unadopted because of a licensing price.
I've earned no real income since March 2025. I'm an unpaid founder, which so far, and perhaps for a long time yet, is the same experience as being an unpaid volunteer: working without financial support to find new, scalable and renovative ways to cut a harm that costs the United States about $3.1 trillion over survivors' lifetimes (2014 dollars, more than $4 trillion today; Peterson and colleagues, 2017) and makes a woman six times more likely to attempt suicide (Davidson and colleagues, 1996). Bringing forward policies made possible by inventions like these should be rewarded, as it would be for a drug that cut the risk of suffering or being accused of sexual assault by 40% or more. Fair licensing should let me earn a living, build a team and fund further experiments against similar harms. When those goals conflict with adoption, adoption wins.
— David Clayman
About
Who is behind this
Mission Zero NY is about every kind of sexual assault in New York, with a special focus on date and acquaintance rape. It starts with campuses because they can run the statewide pilot and measure results, but it is written for every New Yorker who dates, students or not, from senior prom to 81 and beyond, and most of its proposals reach them all. It was written by David Clayman, founder and CEO of Lifesaver Labs PBC, a public benefit corporation that builds consent and emergency-response technology, including the Safeword (Safer Yes) consent system and Neighbor 911. Every proposal where Lifesaver Labs could benefit is marked, and Lifesaver Labs would have no role in evaluating any pilot.
The goal is for New York to take this up as its own, at MissionZeroNY.gov, and then to invite other states through an interstate compact to run the experiment with it, driving rates lower and lower. Legislators, campus leaders, survivors, advocates, researchers and engineers are all welcome to argue with it, improve it and help test it.
david@lifesaverlabs.orgWhy "zero"
We can't reach zero tomorrow. We can get shockingly close.
We know we can't actually reach zero sexual assaults, not immediately and not in the next few years. But if we put our engineering and privacy-engineering talent to this problem, we can drive the number so low it's shocking, and even more shocking when we ask exactly why we didn't try to do this before.