Ratings assume corporations optimize for revenue, not user privacy — even when they claim otherwise. “Metadata” is often more revealing than message content. Interoperability column reflects EU Digital Markets Act (DMA) Article 7 gatekeeper obligations and IETF MIMI / MLS (RFC 9420) protocol alignment. Evaluated as of early 2026.
| App & Owner | Clients A i W M |
Independence from platform / corp |
E2EE / Security as actually deployed |
Metadata Collected known & inferred |
Monetization & Hidden Risks |
Source Openness |
Future‑ proof |
Interoperability DMA / MIMI / XMPP / Matrix |
Verdict |
|---|---|---|---|---|---|---|---|---|---|
| Privacy‑first / Open ecosystems | |||||||||
SignalSignal Foundation (US non‑profit) non‑profitopen source |
AiWM |
High | Very High Default E2EE all chats; Sealed Sender hides metadata |
Donation-funded. No ads, no data sales. Risk: US jurisdiction, donor dependency. |
Full OSS | Med‑High | None planned Below DMA gatekeeper threshold. Explicitly opposes federated interop — argues it creates unverifiable E2EE trust chains with closed-source partners. |
Gold standard for private messaging. Desktop client mature & actively maintained. Only real concern is funding fragility. |
|
Matrix / ElementElement (UK); Matrix.org Foundation federatedopen source |
AiWM No self‑hosting needed |
High* | High Olm/Megolm E2EE; key cross-signing; E2EE on by default in Element |
Enterprise hosting sales. No advertising model. Public servers free; no self-hosting required. |
Full OSS | Very High | Native federated Is an interop standard. Matrix.org Foundation actively lobbied EU to recognise Matrix as a DMA Article 7 compliance path. MLS (RFC 9420) adoption underway for cross-ecosystem E2EE. Real-world federation works today. |
Best long-term bet. Public servers work out of the box — no server to run. Privacy on public server similar to other managed apps. *Self-hosted = Very High independence. |
|
ThreemaThreema GmbH (Switzerland) one‑time purchaseopen source |
AiWM |
High | High NaCl E2EE; no phone/email required; random Threema ID |
Paid app (~€5 one‑time). Zero ads. Swiss privacy law enforced. Fully managed service. |
OSS (2020) | Medium | None Below DMA threshold. Proprietary Threema protocol. No XMPP/Matrix bridge. No federation. |
Very strong privacy; good-quality native clients all platforms. Small user base limits network effect. |
|
SimpleX ChatSimpleX Chat Ltd (UK) — VC‑seed open sourceno user IDs |
AiWM Desktop client newer, less mature |
Very High | Very High Double-ratchet; no user identifiers at all — not even a username |
Free; VC-seeded. No self-hosting needed. Business model unproven. Architecture is uniquely surveillance-resistant. |
Full OSS | Medium | By design: no No stable user identifiers anywhere — structurally incompatible with all standard interop models (DMA, MIMI, XMPP, Matrix) which require addressable users. |
Most metadata-lean design today. Desktop client works but less polished than Signal. Uncertain long-term funding. |
|
SessionOxen Privacy Tech Foundation (AU) open sourcedecentralised |
AiWM |
Very High | High Signal protocol variant; no phone number; onion-routed via Oxen network |
OXEN crypto token funds infrastructure. No ads, no data monetization. Fully managed, no setup. |
Full OSS | Med‑Low | None Decentralised Oxen network. Below DMA threshold. No standard interop protocol. |
Strong anonymity design; good native clients. OXEN token valuation is a sustainability risk. |
|
BriarBriar Project (non‑profit, EU/DE) non‑profitopen sourceP2P / Tor |
AiWM Android only — fails filter |
Very High | Very High P2P via Tor; works over Bluetooth/WiFi with no internet |
Grant-funded, no commercial model. Built for activists & high-risk users. |
Full OSS | Medium | N/A P2P mesh over Tor/BT/WiFi — no server-to-server federation possible by architecture. |
Technically excellent but Android‑only. Excluded from consideration if cross-platform is required. |
|
WireWire Swiss GmbH (CH/DE) open sourceB2B focus |
AiWM |
High | High Default E2EE; MLS protocol for groups. Team/Wire for Business hosted service available. |
Enterprise SaaS. No consumer advertising. Free personal tier exists. Ownership history is a governance concern. |
Full OSS | Medium | MLS-ready First production messaging app to deploy MLS (RFC 9420) — the E2EE layer that MIMI is built on. Best-positioned privacy-focused app for future MIMI federation. |
Solid quality clients all platforms. Consumer focus largely abandoned in favour of B2B. Ownership instability is a red flag. |
|
| Mid‑tier — privacy partially compromised | |||||||||
iMessageApple Inc. (US) corporate |
AiWM Apple‑only — fails filter |
Medium | Partial E2EE Apple-to-Apple; unencrypted SMS fallback; iCloud backup breaks E2EE by default |
Hardware upsell — messages not sold to advertisers. But no Android & no Windows = platform lock-in. |
Closed | Medium | DMA required Apple designated DMA gatekeeper. Interop compliance plan submitted; progress slow and friction-laden. Third-party iMessage access technically available but no Android = interop is still one-sided. |
Excluded — Apple ecosystem only. No Android client exists and none is planned. |
|
ViberRakuten (Japan/US) corporate |
AiWM |
Low | Partial E2EE for 1:1 and group chats; Channels & Communities NOT encrypted |
Ad revenue in public spaces. Rakuten cross-app data sharing. Less aggressive than Meta but no strong privacy guarantees. |
Closed | Low | None Below DMA gatekeeper threshold. Proprietary protocol. No interop roadmap. |
Full cross-platform native clients. Declining user base; partially encrypted but surrounded by ad infrastructure. |
|
Google Messages (RCS)Google / Alphabet (US) corporate |
AiWM Android‑primary — fails filter |
Low | Partial E2EE for 1:1 RCS on Android; group E2EE limited; SMS fallback unencrypted |
Ad-driven Alphabet. Metadata feeds identity graph. Google kills products unpredictably (Allo, Hangouts, +). |
Closed | Low | RCS standard RCS (GSMA) is itself a carrier-level interop standard — works across telcos by design. Different layer from DMA app-level interop. Not a designated DMA gatekeeper for messaging. |
Excluded — Android-first only; no iOS app, no native desktop. Your communication graph is an ad product. |
|
LineLY Corporation (Naver/SoftBank, JP/KR) corporate |
AiWM |
Low | Partial "Letter Sealing" E2EE opt-in; not default everywhere; backups unencrypted |
Sticker market, Line Pay, shopping, advertising. Deep commerce data linkage. Data sovereignty incident 2024. |
Closed | Low | None Below DMA threshold. Dominant in JP/TH/TW but no interop standard. Proprietary protocol. |
Full native clients all platforms. Dominant in Japan/Thailand/Taiwan. Restructuring post-breach adds uncertainty. |
|
| High‑risk — pervasive data collection, authoritarian exposure, or structural deceptions | |||||||||
WhatsAppMeta Platforms (US) corporate |
AiWM |
Very Low | Content only Signal protocol for message content; metadata fully exposed to Meta |
Meta's core business is advertising via behavior profiling. "We don't read messages" is technically true and strategically misleading. |
Closed | Low | DMA API live Third-party chat API deployed 2024 under DMA Article 7. Meta-controlled gateway — technically compliant, deliberately friction-laden. Critics: "compliance as moat." Signal & others declined due to Meta's closed server. |
Polished native clients all platforms. Content encrypted, context fully monetized. The world's largest private social graph extraction operation. |
|
Facebook MessengerMeta Platforms (US) corporate |
AiWM |
Very Low | Low‑Med Default E2EE rolled out 2023‑24 under pressure — years of plaintext history already extracted |
Advertising behemoth. E2EE added under regulatory pressure. Any content sharing analyzed for ad targeting. |
Closed | Low | DMA API live Shares Meta's DMA gatekeeper obligation with WhatsApp. Same gateway API. Meta controls the bridge — no neutral protocol. |
E2EE is damage control, not a privacy redesign. Historical data already extracted and monetized. |
|
TelegramTelegram (Dubai; legal uncertainty post‑Durov arrest) corporate |
AiWM |
Low | Low Default: server-client encryption only (Telegram holds keys). E2EE only in "Secret Chats" — NOT groups, NOT channels |
Marketed as "secure" but NOT E2EE by default. Durov arrest ended independence fiction. TON crypto + Premium + ads = revenue model. |
Client OSS, server closed |
Low | None Not designated a DMA gatekeeper (contested; below threshold or appealing). MTProto is proprietary & closed server-side. No standard interop path. |
Most dangerous gap between privacy perception and reality. Polished clients all platforms. Billions think it is encrypted — it is not by default. |
|
DiscordDiscord Inc. (US) — VC‑backed corporate |
AiWM |
Very Low | None No E2EE whatsoever. All messages readable by Discord and any legal request |
Zero encryption by design. Nitro subscriptions + ads + B2B. Game activity & social behavior = valuable profiling data for youth. |
Closed | Low | None Not a DMA gatekeeper. No interop planned. No E2EE = bridging would expose all content. |
Polished native clients all platforms. Treat every Discord message as readable. Large behavioral dataset with zero encryption protection. |
|
SnapchatSnap Inc. (US) corporate |
AiWM No native desktop |
Very Low | Low E2EE for snaps claimed; Memories & Spotlight server-side; "ephemeral" is a UI metaphor, not cryptographic |
Advertising to teenagers. Biometric data from AR filters. "Disappearing" is a UX feature, not a security guarantee. |
Closed | Low | None Not a DMA gatekeeper. No interop roadmap. Product identity (ephemeral media) incompatible with standard messaging interop. |
Mobile-only product. Among the highest-risk platforms for youth. Biometric + location + social graph + ad targeting on adolescents. |
|
X / Twitter DMsX Corp / xAI (Elon Musk) corporate |
AiWM No native desktop DM client |
Very Low | Very Low E2EE DMs announced (2023) then paused; 2025‑26: partial, premium-only, unverified |
AI data extraction + advertising + arbitrary moderation. DM content explicitly used for Grok AI training. No governance checks on owner. |
Closed | Very Low | Under review DMA gatekeeper status uncertain post-Musk (threshold contested, political complications). No interop implemented. ActivityPub (Mastodon) bridge rumoured but not delivered. |
DMs are an AI training corpus. Worst governance transparency of any major platform. Avoid for anything sensitive. |
|
WeChatTencent (China); CCP access mandated by law corporateauthoritarianism risk |
AiWM |
Very Low | None No E2EE. All messages accessible to Tencent & by law to Chinese state security |
State surveillance infrastructure as a product. Non-Chinese users messaging Chinese accounts trigger monitoring. |
Closed | Very Low | DMA required Designated DMA gatekeeper. Zero compliance implemented. EU enforcement severely limited by Chinese jurisdiction — effectively unenforceable in practice. |
Not just privacy-hostile — actively a state surveillance tool. Necessary for China engagement; dangerous for sensitive communication. |
|
TikTok DMsByteDance (China); legal status volatile US/EU corporateauthoritarianism risk |
AiWM Mobile-first product |
Very Low | Very Low No meaningful E2EE; ByteDance employee backend access documented |
Advertising + algorithmic attention maximization. ByteDance Chinese ownership implies CCP data access obligation. |
Closed | Very Low | DMA required Designated DMA gatekeeper. Under active enforcement for other DMA violations. No messaging interop implemented. Ownership/legal status adds further uncertainty. |
Primary concern is feed algorithm & behavioral extraction. DMs have no privacy protections. Regulatory future uncertain. |
|
Last updated: March 2026 · Ratings are subjective assessments of structural risk based on publicly documented behavior and verified research. Not legal advice.
Status: Legally binding. Compliance deadlines passed. Enforcement ongoing.
The DMA (Regulation (EU) 2022/1925), in force March 2024, designates large platforms as gatekeepers and mandates that they open messaging to third-party interoperability on request. Article 7 applies specifically to number-independent interpersonal communications services.
The DMA mandates that gatekeepers interoperate but says nothing about how. The IETF MIMI working group (More Instant Messaging Interoperability, formed 2022) is producing the application-layer protocol. It is built on top of MLS (Messaging Layer Security, RFC 9420, published 2023) as the shared E2EE primitive. This solves the hardest problem: how do two apps with different encryption stacks exchange end-to-end encrypted messages without one party needing to trust the other's server?
Key tension: messaging apps that use their own proprietary E2EE (Signal Protocol, MTProto, etc.) must either abandon their implementation in favour of MLS, run a translation layer that inevitably weakens the E2EE guarantees at the boundary, or refuse to interop — which is what Signal has done, arguing correctly that cross-protocol E2EE bridges create unauditable trust chains, especially against closed-source gatekeeper servers.
Matrix is the only major ecosystem that is both already federated and actively integrating MLS, positioning it as the most technically coherent DMA compliance path. The Matrix.org Foundation submitted formal position papers to the EU arguing this case.
Status: No binding legislation. Legislative progress stalled as of early 2026.
The US has no equivalent to the DMA. Interoperability has been raised through three separate tracks — dedicated bills, broad platform antitrust legislation, and DOJ/FTC antitrust enforcement — with limited progress in any of them.
The most directly relevant bill: Augmenting Compatibility and Competition by Enabling Service Switching (ACCESS) Act, first introduced 2021 by Senators Mark Warner (D-VA), Mazie Hirono (D-HI), and Richard Blumenthal (D-CT). Reintroduced in modified form in 2023. Key provisions:
Progress: None. Never reached a floor vote in the Senate. The bill was backed by Democrats and opposed by tech industry lobbying. Under the 119th Congress (2025–26, Republican majority), no companion bill has been reintroduced. Senator Warner has continued to advocate but without traction.
Broader platform self-preferencing bill led by Senator Amy Klobuchar (D-MN), co-sponsored bipartisanly (including Sen. Chuck Grassley, R-IA). Passed Senate Judiciary Committee in 2022 — a notable moment. Contained provisions that would have indirectly pressured interoperability by restricting how gatekeepers could preference their own messaging products. Died before a Senate floor vote amid Apple/Google lobbying and WhatsApp E2EE objections. Not reintroduced in meaningful form.
The DOJ filed suit against Apple in March 2024, explicitly citing iMessage as a mechanism for platform lock-in. The complaint argued that Apple deliberately degraded cross-platform messaging (green bubbles, withheld RCS support until late 2023) to make iPhone switching costly. Messaging interoperability is directly framed as a remedy consideration. Case ongoing as of early 2026 — no remedies phase yet. If the government prevails and an interoperability remedy is ordered, it could be more impactful than any bill so far.
Under Lina Khan (2021–24), the FTC discussed data portability and interoperability as structural remedies in platform markets, citing European precedent. Under the current administration (2025–), the FTC has deprioritised this approach. The FTC's Meta enforcement action (seeking divestiture of Instagram/WhatsApp) remains active but is focused on acquisition remedies, not interoperability mandates.
The House has seen less activity than the Senate on interoperability specifically. The Platform Competition and Opportunity Act (Rep. Hakeem Jeffries, D-NY + Rep. Ken Buck, R-CO) addressed acquisitions. Rep. Yvette Clarke's (D-NY) work on algorithmic accountability touched adjacent terrain. No House bill specifically targeting messaging interoperability has advanced past committee.