All articles tagged with #mozilla
Firefox 151 adds desktop support for the Web Serial API, letting websites read from and write to serial devices via JavaScript. Mozilla and Adafruit are collaborating to validate browser-based hardware workflows, enabling direct hardware communication (e.g., microcontrollers) in the browser with less setup and no extra software for many projects.
Firefox 153 Nightly now enables a redesigned Settings UI by default, offering clearer organization, easier navigation, and improved descriptions while keeping AI controls. Mozilla hopes the redesign will eventually reach the Firefox 152 release channel, with more details on the Nightly Blog.
Mozilla says Anthropic's Mythos Preview helped uncover hundreds of security flaws, driving 423 Firefox bug fixes in April (271 tied to Mythos) including issues undetected by fuzzers and one lingering for 20 years; January had 25 fixes and March 76, signaling a dramatic jump in AI-assisted discovery as Mythos remains restricted to a few partners.
Mozilla says Mythos, guided by a custom harness, identified 271 Firefox vulnerabilities in two months with 'almost no false positives,' using the same testing pipeline as human developers. Of these, 180 were sec-high, 80 sec-moderate, and 11 sec-low; 12 Bugzilla reports have been released so far. Critics question cherry-picking and CVE designations, but Mozilla argues the harness plus a second LLM verifier makes AI-assisted discovery scalable and reliable, not marketing.
Mozilla says Claude Mythos helped identify and patch 271 Firefox vulnerabilities, but the model didn’t find bugs beyond what human researchers would, and the feature can be turned off by users.
Mozilla's Firefox 150 release uses Anthropic Mythos Preview to identify and fix 271 vulnerabilities, marking a major AI-assisted bug-hunting milestone; Mozilla cautions that the industry will have to go through a difficult transition to secure software with these tools, especially for open-source projects, while anticipating that such capabilities will become common for both defenders and attackers.
Mozilla announced Thunderbolt, an open-source, self-hosted AI client for enterprises that lets organizations run AI with their chosen models, connect to data and workflows via Haystack, Model Context Protocol (MCP) and Agent Client Protocol (ACP), automate tasks, and use native apps across Windows, macOS, Linux, iOS, and Android, with optional end-to-end encryption and device-level controls; the code is MPL 2.0-licensed on GitHub.
Mozilla’s Firefox 149 will include a free built-in VPN (50 GB/month in the US, France, Germany and the UK) starting March 24, plus two new features—Split View and Tab Notes—while also introducing a new mascot named Kit, all as part of Mozilla’s push for privacy and open web standards.
Anthropic’s Claude Opus 4.6 reportedly scanned Firefox for bugs, finding over 100 issues in two weeks, including 14 high-severity flaws. Mozilla validated and fixed many of these findings and expanded collaboration to apply the approach across more of the browser codebase, highlighting AI-assisted security analysis as a powerful addition to traditional fuzzing and static analysis.
Anthropic’s Claude Opus 4.6 identified 22 Firefox vulnerabilities (14 high, 7 moderate, 1 low) during a two-week security review with Mozilla, with most fixes shipped in Firefox 148. The AI scanned about 6,000 C++ files, produced 112 reports, and in testing could automatically develop crude exploits for two issues (including CVE-2026-2796, a JIT miscompilation in WebAssembly), though only in sandbox-stripped environments; a task verifier helped determine exploit viability. Mozilla says AI-assisted analysis uncovered roughly 90 additional bugs and underscored AI as a powerful complement to security engineering, while noting patches remain under active refinement.
Mozilla's Firefox 148 introduces centralized AI controls, including a 'Block AI enhancements' kill switch that can disable all AI features across the browser—covering translation, PDF alt-text generation, AI-driven tab grouping, link previews, and the built-in AI chatbot sidebar—emphasizing privacy and security with Nightly early access for enterprise testing.
Mozilla will add an AI controls panel in Firefox 148 featuring a 'Block AI enhancements' toggle that lets users block all current and future AI features or selectively enable five tools (translations, image alt text for PDFs, AI‑assisted tab naming, link previews, and a chatbot sidebar). The controls begin in Nightly on Feb 24 and roll out to all desktop users later in February, reflecting Mozilla’s emphasis on user choice and persistence of preferences across updates.
Mozilla says Firefox 148, rolling out Feb. 24, will add per-feature AI controls (translations, PDF alt text, AI-powered tab grouping, link previews, and the AI chatbot in the sidebar) plus a global Block AI Enhancements toggle, allowing users to disable all current and future AI features or choose which to use.
Mozilla's Firefox plans to include an option to disable all AI features, dubbed the 'AI kill switch,' in response to user backlash against AI integration in browsers. Despite assurances of user control, concerns remain about the true opt-in nature of AI features, with some users skeptical about Mozilla's commitment to user choice. Meanwhile, competitors like Vivaldi emphasize maintaining user autonomy and privacy, rejecting passive AI integration.
Waterfox, a Firefox fork, is taking a stand against AI integration, explicitly refusing to include large language models (LLMs) in its browser, in contrast to Mozilla's plans for a more AI-enabled Firefox ecosystem. Mozilla has responded by promising an opt-in AI feature with a kill switch, but Waterfox's stance appeals to users seeking an AI-free browsing experience. Waterfox continues to cater to users who value privacy, customization, and legacy add-ons, positioning itself as a preferred alternative amidst Mozilla's evolving AI ambitions.