As has been the case for several years, Google revealed the conference's dates for 2026 after enough folks completed a puzzle on the I/O website. This year's puzzle has multiple "builds" to play through, all of which use Gemini. They start with a mini-golf game in which a virtual caddy that's powered by Gemini offers some of the most anodyne advice imaginable.
The firm's Chief Operating Officer Marshall Beard, Chief Financial Officer Dan Chen and Chief Legal Officer Tyler Meade are all leaving the company effective immediately, New York-based Gemini said in a filing on Tuesday. Beard, who has also stepped down from Gemini's board of directors, has not resigned as a result of any disagreement with the firm, it added. Gemini does not intend to appoint a successor to Beard at this time, it said. Instead, President Cameron Winklevoss will assume several of his duties.
On Thursday, Google announced that "commercially motivated" actors have attempted to clone knowledge from its Gemini AI chatbot by simply prompting it. One adversarial session reportedly prompted the model more than 100,000 times across various non-English languages, collecting responses ostensibly to train a cheaper copycat. Google published the findings in what amounts to a quarterly self-assessment of threats to its own products that frames the company as the victim and the hero, which is not unusual in these self-authored assessments.
Google on Thursday said it observed the North Korea-linked threat actor known as UNC2970 using its generative artificial intelligence (AI) model Gemini to conduct reconnaissance on its targets, as various hacking groups continue to weaponize the tool for accelerating various phases of the cyber attack life cycle, enabling information operations, and even conducting model extraction attacks. "The group used Gemini to synthesize OSINT and profile high-value targets to support campaign planning and reconnaissance,"
A Chinese government hacking group that has been sanctioned for targeting America's critical infrastructure used Google's AI chatbot, Gemini, to auto-analyze vulnerabilities and plan cyberattacks against US organizations, the company says. While there's no indication that any of these attacks were successful, "APT groups like this continue to experiment with adopting AI to support semi-autonomous offensive operations," Google Threat Intelligence Group chief analyst John Hultquist told The Register. "We anticipate that China-based actors in particular will continue to build agentic approaches for cyber offensive scale."
Google announced on Tuesday that the public preview of Fitbit's AI personal health coach is rolling out to iOS users in the U.S. as well as both iOS and Android users in the UK, Canada, Australia, New Zealand, and Singapore. It initially launched in October, when it was only available in the U.S. on Android devices. To try out the feature, users need to have an active Fitbit Premium subscription and use a Google account to sign into the Fitbit app.
Google is expanding its push into AI-powered learning by adding full-length practice tests in Gemini for the Joint Entrance Exam (JEE), India's nationwide engineering exam used to shortlist candidates for the country's top technical institutes and taken by millions of students each year. Google said students can take full-length mock exams for the JEE within Gemini, which will offer questions based on vetted content from Indian education firms PhysicsWallah and Careers360.
Perhaps sensing that feeding more data into Gemini would give many people the creeps, Google's announcement explains at great length how the company has approached privacy in Personal Intelligence. Google isn't getting any new information about you-your photos, email, and search behaviors are already stored on Google's servers, so "you don't have to send sensitive data elsewhere to start personalizing your experience."
Google's latest addition to its NotebookLM artificial intelligence research platform is a feature called Data Tables. The tool can collect and synthesize information across multiple sources into a chart that can be exported to Google Sheets. All Pro and Ultra users will have access to the feature today, and Data Tables will roll out to all users over the coming weeks.
Gemini will scan the video's visuals and audio for Google's proprietary watermark called SynthID. The response will be more than a yes or no, Google says. Gemini will point out specific times when the watermark appears in the video or audio. The company rolled out this capability for images in November, also limited to images made or edited with Google AI.