OpenAI just banned accounts linked to five covert influence operations originating from Russia, China, Iran, and Israel. Malicious actors used OpenAI's models to churn out multilingual comments, write political articles, and debug bot code across platforms like X and Telegram. Why it matters: Despite using generative AI to scale up spam, none of these networks managed to break through or gain meaningful traction with real audiences.
Elon Musk and Meta's Yann LeCun just clashed in an unfiltered spat on X over scientific publishing and AI hype. After LeCun slammed Musk's apocalyptic claims and wild deadlines, Musk questioned LeCun's recent research credentials. Why it matters: The clash exposes the widening rift between open academic science and secretive, hype-driven commercial labs.
Former OpenAI board member Helen Toner just revealed that the board had no prior knowledge of ChatGPT's November 2022 release and only found out through Twitter. She also claimed CEO Sam Altman concealed his ownership of the OpenAI startup fund and misled the board on safety procedures. Why it matters: It gives the clearest picture yet of the communications breakdown and lack of transparency that led to Altman's brief firing last fall.
Former OpenAI Superalignment co-lead Jan Leike just joined rival Anthropic to head up a new superalignment team, weeks after publicly criticizing OpenAI's safety approach. Leike's team will focus on scalable oversight, weak-to-strong generalization, and automated alignment research. The high-profile hire bolsters Anthropic's safety-first reputation right as OpenAI faces scrutiny over dissolving its own team.
OpenAI just formed a new Safety and Security Committee to oversee critical safeguards across its projects, led by CEO Sam Altman and board chair Bret Taylor. The company also confirmed it has begun training its next frontier model, with the committee tasked with reviewing safety processes over the next 90 days before publicly sharing adopted recommendations.
Character.ai is reportedly exploring research partnerships with Meta and xAI to collaborate on pre-training and AI model development. Sources say the talks center on joint research rather than an acquisition, with no deal signed yet. Facing rising antitrust scrutiny over buyouts, tech giants are increasingly relying on strategic tie-ups to tap frontier tech.
OpenAI just sent an internal memo releasing former employees from controversial non-disparagement agreements that previously tied their silence to their vested equity. The company confirmed it has never canceled and will not cancel vested shares over unsigned exit paperwork, apologizing that the language did not reflect its values.
Google's new search feature AI Overviews is going viral for the wrong reasons. The tool instructed a user to add non-toxic glue to pizza sauce to keep cheese sticking, seemingly falling for an 11-year-old Reddit joke. It also mixed up the Tower of London with Big Ben. Google claims these hallucinations are extremely rare queries, but it highlights how messy search summaries can get.
Mark Zuckerberg just assembled an advisory council to guide Meta’s AI roadmap, tapping Stripe CEO Patrick Collison, Shopify chief Tobi Lütke, and former GitHub CEO Nat Friedman. The unpaid group sits outside the board of directors to offer pure strategic advice on Meta's tech push.
World leaders and 16 frontier AI companies just signed landmark safety commitments at the AI Seoul Summit. Ten nations and the EU agreed to launch an international safety research network, while tech giants including OpenAI, Google, and China's Zhipu.ai pledged to halt the deployment of any model if risks breach predefined thresholds. Why it matters: It is the first time leading AI labs across the US, Europe, and China have formally aligned on explicit red lines.
The UK data watchdog just launched an inquiry into Microsoft's new Recall feature, which takes screenshots of Copilot+ PC screens every few seconds to build a searchable history. While Microsoft insists all data stays encrypted locally and is strictly opt-in, privacy experts warn that capturing passwords and sensitive records without filtering poses a serious security risk.
Scarlett Johansson just called out OpenAI for releasing a ChatGPT voice that sounded eerily like her own, forcing the company to pause the controversial "Sky" option. Johansson revealed she explicitly rejected Sam Altman's pitch to voice the system months ago, hiring legal counsel after hearing the viral GPT-4o demo. OpenAI claims the voice belongs to another actress, but pausing the rollout shows Hollywood's pushback is landing.
Microsoft just unveiled Recall for Copilot+ PCs at Build 2024, using AI to fix Windows' notoriously clunky search. Running over 40 local multimodal small models, it lets you search virtually anything you've seen—from past emails to specific meeting slides—using vague contextual clues. Everything runs offline on-device, rolling out on June 18.
OpenAI just detailed the five-month casting process behind ChatGPT's voices while pausing the controversial Sky voice after comparisons to Scarlett Johansson. CEO Sam Altman insisted the voice belongs to a different actress and was never meant to mimic Johansson, despite reaching out to her twice. Why it matters: As voice AI gets eerily realistic, celebrity soundalikes are rapidly turning into a major legal and PR battlefield.
Slack just sparked massive community backlash after users discovered it automatically opts customer data into training global AI models—and forces them to send an email to opt out. Slack quickly clarified that its paid Slack AI add-on does not train on customer data, but the confusing terms and sneaky defaults have left users deeply frustrated.
Sony Music Group just fired off warning letters to over 700 AI companies and streaming platforms, outright banning unauthorized scraping of its recordings, lyrics, and metadata for AI training. The music giant demanded recipients either confirm they haven't used its catalog or disclose full details on what went into their models. Why it matters: The music industry is drawing a hard legal line before unlicensed AI training becomes the default.
OpenAI's Superalignment team leaders Ilya Sutskever and Jan Leike have both resigned, sparking concerns about the company's commitment to AI safety. Several other safety researchers who didn't back Sam Altman during the boardroom coup have also exited. Why it matters: When the very people tasked with keeping superintelligence in check walk out, questions naturally mount over who is steering the guardrails.
Voice actors just slapped AI voice-cloning startup LOVO with a proposed class action lawsuit in federal court. The plaintiffs allege LOVO duped them on freelance platforms into recording voice samples under the guise of "academic research," only to repackage them as flagship paid AI voices. It marks a critical legal showdown over unauthorized training data and voice theft.
OpenAI co-founder and chief scientist Ilya Sutskever just officially resigned, marking the end of a dramatic saga that began with last November's boardroom shakeup. Research director Jakub Pachocki is stepping into the chief scientist role, while superalignment co-lead Jan Leike is also leaving. Sutskever teased that he is shifting focus to an undisclosed project that is personally meaningful.
Cisco just detailed its AI-native Hypershield architecture at the RSA Conference, aiming to tackle long-standing segmentation and patching bottlenecks. Powered by AI, hardware acceleration, and Extended Berkeley Packet Filter for deep kernel visibility, the system can apply compensating controls across enforcement points in 9 to 15 minutes. It buys critical response time when traditional patches typically take 45 days.
