diff --git a/proposals/3951-intentional-mentions.md b/proposals/3951-intentional-mentions.md
new file mode 100644
index 00000000000..38ff13fe83c
--- /dev/null
+++ b/proposals/3951-intentional-mentions.md
@@ -0,0 +1,268 @@
+# MSC3951: Intentional Mentions
+
+Mentioning other users on Matrix is difficult -- it is not possible to know if
+[mentioning a user by display name or Matrix ID](https://github.com/matrix-org/matrix-spec/issues/353)
+will count as a mention, but is also too easy to mistakenly mention a user.
+
+(Note that throughout this proposal "mention" is considered equivalent to a "ping"
+or highlight notification.)
+
+Some situations that result in unintentional mentions include:
+
+* Replying to a message will re-issue pings from the initial message due to
+  [fallback replies](https://spec.matrix.org/v1.5/client-server-api/#fallbacks-for-rich-replies). [^1]
+* Each time a message is edited the new version will be re-evaluated for mentions.
+* Mentions occurring [in spoiler contents](https://github.com/matrix-org/matrix-spec/issues/16)
+  or [code blocks](https://github.com/matrix-org/matrix-spec/issues/15) are
+  evaluated.
+* If the [localpart of your Matrix ID is a common word](https://github.com/matrix-org/matrix-spec-proposals/issues/3011)
+  then the push rule matching usernames (`.m.rule.contains_user_name`) matches
+  too often (e.g. Travis CI matching if your Matrix ID is `@travis:example.com`).
+* If the [localpart or display name of your Matrix ID matches the hostname](https://github.com/matrix-org/matrix-spec-proposals/issues/2735)
+  (e.g. `@example:example.com` receives notifications whenever `@foo:example.com`
+  is replied to).
+
+As a sender you do not know if including the user's display name or Matrix ID would
+even be interpreting as a mention (see [issue 353](https://github.com/matrix-org/matrix-spec/issues/353)).
+
+This also results in some unexpected behavior and bugs:
+
+* Matrix users use "leetspeak" when sending messages to avoid mentions (e.g.
+  referring to M4tthew instead of Matthew).
+* Matrix users will append emoji or other unique text in their display names to
+  avoid unintentional pings.
+* It is impossible to ping one out of multiple people with the same localpart
+  (or display name).
+* Since the relation between `body` and `formatted_body` is ill-defined and
+  ["pills" are converted to display names](https://github.com/matrix-org/matrix-spec/issues/714),
+  this can result in missed messages.
+* Bridges [must use display names](https://github.com/matrix-org/matrix-spec/issues/353#issuecomment-1055809364)
+  as a trick to get pings to work.
+* If a user changes their display name in a room than whether or not they are
+  mentioned changes unless you use historical display names to process push rules.
+  (TODO I think there's an issue about this?)
+
+## Background
+
+Mentions are powered by two of the default push rules that search an event's
+`content.body` field for the current user's display name
+([`.m.rule.contains_display_name`](https://spec.matrix.org/v1.5/client-server-api/#default-override-rules))
+or the localpart of their Matrix ID ([`.m.rule.contains_user_name`](https://spec.matrix.org/v1.5/client-server-api/#default-content-rules)).
+
+There's also a [section about "user and room mentions"](https://spec.matrix.org/v1.5/client-server-api/#user-and-room-mentions)
+which defines that messages which mention the current user in the `formatted_body`
+of the message should be colored differently:
+
+> If the current user is mentioned in a message (either by a mention as defined
+> in this module or by a push rule), the client should show that mention differently
+> from other mentions, such as by using a red background color to signify to the
+> user that they were mentioned.
+
+## Proposal
+
+Instead of searching a message's `body` for the user's display name or the localpart
+of their Matrix ID, it is proposed to use a field specific to mentions,[^2] and
+augment the push rules to search for the current user's Matrix ID.
+
+### New event field
+
+A new `mentions` field is added to the event content, it is an array of strings
+consistent of Matrix IDs.
+
+To limit the potential for abuse, only the first 10 items in the array should be
+considered. It is recommended that homeservers reject locally created events with
+more than 10 mentions with an error with a status code of `400` and an errcode of
+`M_INVALID_PARAM`.
+
+Clients should add a Matrix ID to this array whenever composing a message which
+includes an intentional [user mention](https://spec.matrix.org/v1.5/client-server-api/#user-and-room-mentions)
+(often called a "pill"). Clients may also add them at other times when it is
+obvious the user intends to explicitly mention a user.
+
+The `mentions` field is part of the plaintext event body and should be encrypted
+into the ciphertext for encrypted events.
+
+### New push rules
+
+A new push rule condition and a new default push rule will be added:
+
+```json
+{
+    "rule_id": ".m.rule.user_is_mentioned",
+    "default": true,
+    "enabled": true,
+    "conditions": [
+        {
+            "kind": "is_mention"
+        }
+    ],
+    "actions": [
+        "notify",
+        {
+            "set_tweak": "sound",
+            "value": "default"
+        },
+        {
+            "set_tweak": "highlight"
+        }
+    ]
+}
+```
+
+The `is_mention` push condition is defined as[^3]:
+
+> This matches messages where the `content.mentions` is an array containing the
+> owner’s Matrix ID in the first 10 entries. This condition has no parameters.
+> If the `mentions` key is absent or not an array then the rule does not match;
+> any array entries which are not strings are ignored, but count against the limit.
+
+An example matching event is provided below:
+
+```json
+{
+    "content": {
+        "body": "This is an example mention @alice:example.org",
+        "format": "org.matrix.custom.html",
+        "formatted_body": "<b>This is an example mention</b> <a href='https://matrix.to/#/@alice:example.org'>Alice</a>",
+        "msgtype": "m.text",
+        "mentions": ["@alice:example.org"]
+    },
+    "event_id": "$143273582443PhrSn:example.org",
+    "origin_server_ts": 1432735824653,
+    "room_id": "!somewhere:over.the.rainbow",
+    "sender": "@example:example.org",
+    "type": "m.room.message",
+    "unsigned": {
+        "age": 1234
+    }
+}
+```
+
+### Backwards compatibility
+
+The the [`.m.rule.contains_display_name`](https://spec.matrix.org/v1.5/client-server-api/#default-override-rules)
+and [`.m.rule.contains_user_name`](https://spec.matrix.org/v1.5/client-server-api/#default-content-rules)
+push rules are both deprecated. To avoid the unintentional mentions they are both
+modified to only apply when the `mentions` field is missing. As this is for
+backwards-compatibility it is not implemented using a generic mechanism, but
+behavior specific to those push rules with those IDs.
+
+While this is being deployed there will be a mismatch for legacy clients which
+implement the deprecated `.m.rule.contains_display_name` and `.m.rule.contains_user_name`
+push rules locally while the `.m.rule.user_is_mentioned` push rule is used on
+the homeserver; as messages which
+[mention the user should also include the user's localpart](https://spec.matrix.org/v1.5/client-server-api/#user-and-room-mentions)
+in the message `body` it is likely that the `.m.rule.contains_user_name`
+will also match. It is postulated that this will not cause issues in most cases.
+
+## Potential issues
+
+### Abuse potential
+
+This proposal makes it trivial to "hide" mentions since it does not require the
+mentioned Matrix IDs to be part of the displayed text. This is not seen as
+worse than what is possible today.
+
+From discussions and research while writing this MSC there are some benefits to
+using a separate field for mentions:
+
+* The number of mentions is trivially limited.
+* Various forms of "mention bombing" are no longer possible.
+* It is simpler to collect data on how many users are being mentioned (without
+  having to process the textual `body` for every user's display name and local
+  part).
+
+Nonetheless, the conversations did result in some ideas to combat the potential
+for abuse, many of which apply regardless of whether this proposal is implemented.
+
+Clients could expose *why* an event has caused a notification and give users inline
+tools to combat potential abuse. For example, a client might show a tooltip:
+
+> The sender of the message (`@alice:example.org`) mentioned you in this event.
+>
+> Block `@alice:example.org` from sending you messages? `[Yes]` `[No]`
+
+It could also be worth exposing user interface for moderators to see messages
+which mention many users.
+
+A future MSC might wish to explore features for trusted contacts or soft-ignores
+to give users more control over who can generate notifications.
+
+Overall the proposal does not seem to increase the potential for malicious behavior.
+
+## Alternatives
+
+### Prior proposals
+
+There's a few prior proposals which tackle subsets of the above problem:
+
+* [MSC2463](https://github.com/matrix-org/matrix-spec-proposals/pull/2463):
+  excludes searching inside a Matrix ID for localparts / display names of other
+  users.
+* [MSC3517](https://github.com/matrix-org/matrix-spec-proposals/pull/3517):
+  searches for Matrix IDs (instead of display name / localpart) and only searches
+  specific event types & fields.
+* [Custom push rules](https://o.librepush.net/aux/matrix_reitools/pill_mention_rules.html)
+  to search for matrix.to links instead of display name / localpart.
+
+  <summary>
+
+  This generates a new push rule to replace `.m.rule.contains_display_name` and
+  `.m.rule.contains_user_name`:
+
+  ```json
+  {
+     "conditions": [
+       {
+           "kind": "event_match",
+           "key": "content.formatted_body",
+           "pattern": "*https://matrix.to/#/@alice:example.org*"
+       }
+     ],
+     "actions" : [
+       "notify",
+       {
+         "set_tweak": "sound",
+         "value": "default"
+       },
+       {
+         "set_tweak": "highlight"
+       }
+     ]
+  }
+  ```
+
+  </summary>
+
+### Room version for backwards compatibility
+
+Alternative backwards compatibility suggestions included using a new room version,
+similar to [MSC3932](https://github.com/matrix-org/matrix-spec-proposals/pull/3932)
+for extensible events. This does not seem like a good fit since room versions are
+not usually interested in non-state events. It would additionally require a stable
+room version before use, which would unnecessarily delay usage.
+
+## Security considerations
+
+None foreseen.
+
+## Unstable prefix
+
+During development the `.org.matrix.msc3951.is_user_mentioned` push rule will be
+used. If a client sees this rule available it should apply the custom logic discussed
+in the [backwards compatibility](#backwards-compatibility) section.
+
+## Dependencies
+
+N/A
+
+[^1]: This MSC does not attempt to solve this problem, but [MSC2781](https://github.com/matrix-org/matrix-spec-proposals/pull/2781)
+proposes removing reply fallbacks which would solve it. Although as noted in
+[MSC3676](https://github.com/matrix-org/matrix-spec-proposals/pull/3676) this
+may require landing [MSC3664](https://github.com/matrix-org/matrix-doc/pull/3664)
+in order to receive notifications of replies.
+
+[^2]: As proposed in [issue 353](https://github.com/matrix-org/matrix-spec/issues/353).
+
+[^3]: A new push condition is necessary since none of the current push conditions
+(e.g. `event_match`) can process arrays.