Telegram Widgets Spark RCE Firestorm Days After Launch

Telegram’s newest feature lasted barely 48 hours before security researchers proved it could be weaponised to run arbitrary code on Android devices. The fallout has included emergency patches, public clashes between Telegram’s founder and the security community, and warnings from Kaspersky about possible malware distribution — all in the span of a single week.

What Are Telegram Widgets?

On November 9, 2026, Telegram rolled out HTML widgets as part of Bot API 10.4. The feature allows bots to send interactive, richly formatted mini-applications directly inside chat messages — countdown timers, styled cards, embedded buttons — without requiring users to open a separate Mini App. Telegram founder Pavel Durov announced the feature with characteristic enthusiasm, declaring that any Telegram message could now become an app.

Widgets represented a significant expansion of Telegram’s bot ecosystem. They also arrived alongside integrated Gram wallet payments through the TON blockchain, with Durov reporting that over one million self-custody wallets were activated within 48 hours of launch. The combination of executable content embedded in chat messages and financial functionality immediately caught the attention of the security research community.

V12’s RCE Proof of Concept

On November 11, security firm V12 published a video demonstration that reworked Durov’s own marketing tagline into a warning: “Any Telegram message can now become RCE.” In the proof-of-concept video, a user interacts with a bot by tapping the standard /start command. The bot responds with a widget message, and seconds later, the phone’s Calculator app opens on its own — the classic demonstration that an outsider has achieved remote code execution on the device.

V12 did not publish the technical details of the exploit or release any code, but the video was clear and unambiguous. The demonstration showed that the new widget rendering pipeline in Telegram for Android contained a flaw serious enough to allow a malicious bot to break out of the expected sandbox and execute commands at the device level.

Telegram responded quickly. Within roughly two hours of V12’s disclosure, the company confirmed the issue in Telegram for Android had been patched the previous day and that the fix had been deployed to all users. A client-side update, Telegram for Android 13.2.2, was subsequently pushed to reinforce the server-side mitigation.

Durov’s Dare and the Disputed Severity

Rather than issuing a conventional security advisory, Durov took to his Telegram channel on November 13 and challenged the entire hacking community to break into his personal account. He argued the flaw “doesn’t exist” as described, claiming the exploit required a third-party service to be active first, which would sharply narrow the pool of potential victims. In Durov’s framing, the issue was not a true remote code execution vulnerability because user interaction with a bot was required.

The security community saw it differently. While clicking /start on a bot is technically user-initiated, most Telegram users do so routinely and without suspicion. The gap between “requires one tap on a bot message” and “fully remote” is thin in practice, particularly when the attack can be delivered through any bot in a user’s chat list. Durov’s public dare drew criticism from researchers who felt it minimised a legitimate risk and could discourage responsible disclosure.

NebuSec Claims Sandbox Bypass

The story escalated on November 14 when NebuSec, a security firm that had previously discovered a separate voice-call RCE flaw in Telegram for Android in October, published a new video demo. NebuSec claimed its exploit defeats seccomp, the Linux system-call filter that Telegram had recently added to sandbox the process responsible for rendering widget content. If accurate, the claim would mean Telegram’s freshly deployed mitigation was already insufficient.

NebuSec noted it ran the exploit against Telegram’s test server, explaining that Telegram had restricted widget functionality to whitelisted bots on the production service as an additional safeguard. No CVE, affected build number, or technical write-up accompanied the demo. Telegram had not publicly responded to NebuSec’s claim at the time of the demonstration.

This was the second public widget code-execution demonstration in under a week, arriving from an independent research group with an established track record of finding critical Telegram vulnerabilities.

Kaspersky Suspects Active Exploitation

Adding another layer of urgency, Kaspersky flagged suspected exploitation of Telegram’s widget feature to deliver WebRAT, a remote access trojan. SecurityWeek reported that the antivirus vendor believes attackers may already be taking advantage of the new widget functionality to push malware, though the extent of in-the-wild exploitation remains unconfirmed at this stage. The concern is not theoretical: Telegram’s massive user base, combined with the novelty of widgets — which most users are encountering for the first time — creates ideal conditions for social engineering attacks disguised as legitimate bot interactions.

The Crypto Dimension

Cryptocurrency exchange Bybit issued a direct warning to its users following the widget vulnerability disclosures. The exchange cautioned that a compromised phone could put all applications at risk, including crypto wallets. This is especially pertinent given that Telegram’s widgets are designed to carry Gram payments through the TON blockchain’s self-custody wallets — wallets where no company can reverse a fraudulent transfer once it is confirmed. An attacker who achieves code execution on a device running both Telegram and a crypto wallet has the potential to drain funds irreversibly.

A Pattern of Telegram Android Vulnerabilities

The widget RCE is not an isolated incident. NebuSec’s October disclosure involved a separate vulnerability in Telegram’s voice-call functionality that could achieve RCE when a victim answered a call. That flaw, which affected Android versions dating back to 2022, was patched in Telegram for Android 12.10.4. Earlier in the year, the Zero Day Initiative (ZDI) disclosed an unpatched vulnerability in Telegram’s animated sticker rendering library rlottie, rated CVSS 7.0, which could allow code execution through malicious files.

The recurring pattern points to a broader challenge: Telegram’s rapid feature development pace may be outstripping its security review capacity. Each new interaction surface — voice calls, stickers, widgets — introduces rendering and parsing code that must handle untrusted input from potentially malicious sources. When these features ship with critical flaws that are discovered within days, it suggests security testing is not keeping pace with the development roadmap.

What You Should Do

The immediate advice is straightforward. Update Telegram for Android to the latest version available on the official app store — do not delay. Avoid interacting with unfamiliar bots, particularly those that send widget messages or prompt you to tap buttons with unfamiliar content. If your organisation uses Telegram, push the update through your MDM platform. Treat widget messages from bots you do not recognise as potentially hostile.

Beyond the immediate fix, this incident is a reminder that feature-rich messaging platforms carry an expanding attack surface. When a new capability is announced, the prudent approach is to wait for the security community’s assessment before adopting it in environments where compromise carries serious consequences — especially where cryptocurrency or sensitive communications are involved.

Sources: SecurityWeek, Cryptobriefing, BeInCrypto, Dark Web Intelligence, Kod.ru, DEV Community analysis

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.