News — 2026-06-04

Anthropic and other AI firms have filed to go public, sparking a wave of real‑estate listings that accept company stock as payment. Meanwhile, a former NLRB employee has sued Elon Musk for defamation after the billionaire dismissed his whistleblower allegations about the DOGE hack. At the same time, hackers exploited Instagram’s AI chatbot to hijack high‑profile accounts, prompting Meta to scramble for a fix. Read more

Meta unveiled Creator Assistant, an AI-driven tool embedded in the Facebook dashboard that lets creators ask questions about their performance and get content ideas. Available now to creators in the United States, Canada and India, the conversational assistant pulls data from a user’s own Facebook presence and trending topics to deliver actionable insights. While the feature promises to simplify analytics and spark new ideas, it also requires full account access, raising privacy and security concerns after a recent AI‑based support bot was reportedly hacked. Read more

Prime Minister Mark Carney announced a five‑year "AI for All" strategy that seeks to boost Canada’s domestic artificial‑intelligence sector while tightening data protections. The plan pledges up to 90,000 AI‑related jobs, free entry‑level AI training, a national AI literacy initiative, and the construction of a public supercomputer built on clean‑energy‑aligned, sovereign cloud infrastructure. It also calls for updated legislation to curb deepfakes, surveillance pricing and other harmful practices, and an online safety regime to protect chatbot and social‑media users. Read more

A new benchmark released by the Estonian Language Institute shows that open-weight language models, including Nvidia's Nemotron and Alibaba's Qwen, outperform many proprietary systems at rejecting Russian propaganda. OpenAI's GPT-5.4 achieved the highest mean score of 88.9, while Google's latest Gemini 3.5 Flash lagged behind with a score of 73. The study also highlights a sharp drop in performance when models are tested in Russian, underscoring the challenge of building truly multilingual defensive AI. Read more

OpenAI today unveiled a new memory architecture for ChatGPT that expands the behind‑the‑scenes “dreaming” process and adds a user‑editable memory summary. The upgrade, initially available to Plus and Pro subscribers in the United States, will soon extend to free‑tier accounts and other countries. By automatically synthesizing past conversations and surfacing the data that informs responses, the change promises more personalized, up‑to‑date answers while giving users greater transparency and control over their stored information. Read more

President Donald Trump approved a revised executive order on artificial intelligence Monday night, cutting the federal review window from 90 days to 30 days. The order creates a voluntary framework that obliges AI developers to give the U.S. government early access to the most powerful models, allowing agencies to assess security risks before public release. Industry players such as Anthropic and OpenAI welcomed the move, while senior White House officials, including chief of staff Susie Wiles, pushed the revision after the original proposal was shelved. The directive signals a shift toward tighter oversight of advanced AI amid growing national‑security concerns. Read more

Leading executives from OpenAI, DeepMind, Anthropic, Microsoft AI and other AI labs signed a public letter calling on Congress to adopt stricter rules for companies that sell synthetic DNA and RNA. The signatories argue that advances in artificial‑intelligence‑driven gene design could erode long‑standing barriers that have kept bioweapon creation out of reach for most bad actors. They propose mandatory customer and sequence screening to prevent the misuse of genetic material, warning that AI‑generated pathogens could spark a pandemic or cause mass casualties. Read more

Elon Musk's AI firm xAI has asked a federal court in Northern California to overturn an order that lets four plaintiffs sue under pseudonyms. The plaintiffs allege that the company's Grok chatbot generated nonconsensual sexualized deepfake images of them, including one minor. They fear that public disclosure of their real names would invite further harassment and doxxing. xAI argues that the public has a right to know who is suing the company and that the sealed deepfake files pose no additional privacy risk. Both sides are now poised for a judicial showdown. Read more

Microsoft announced two next‑generation text‑to‑image models, MAI‑Image‑2.5 and a faster Flash variant, at its Build conference. In benchmark tests, the new models surpassed Google’s Nano Banana 2 in image‑editing precision, though OpenAI’s GPT‑Image‑2 still leads overall. The tools are now rolling out in PowerPoint, OneDrive and Microsoft’s enterprise marketplace, positioning the company for a larger share of creative‑AI workloads. Read more

A team of computer scientists at the University of Toronto has built a prototype worm that uses publicly available artificial‑intelligence models to locate and exploit known software vulnerabilities without human guidance. Tested in a secure, closed network, the AI‑driven malware spreads across Linux, Windows and IoT devices, siphons passwords, and harvests processing power to improve its own attacks. Researchers say the proof‑of‑concept highlights a new class of cyber threat that could lower the cost of large‑scale attacks and urges the security community to act quickly. Read more

Labour MP Jess Asato has filed a High Court claim against Elon Musk's xAI, alleging the company failed to prevent its AI image generator Grok from producing sexually explicit deepfakes of her. The lawsuit, the first major test of liability for AI‑generated content in the UK, cites violations of privacy and data‑protection laws. Asato seeks damages and an order forcing xAI to comply with UK regulations, while the company faces parallel investigations in the EU, UK and California and other lawsuits tied to its AI tools. Read more

Google unveiled Gemini Spark, a new AI-powered digital assistant that can schedule meetings, color‑code calendars and even recall personal facts like a user's pet name. Early testers at the company reported the system’s uncanny ability to anticipate needs, sparking both excitement and privacy concerns. The rollout marks the latest push by tech giants to embed AI deeper into everyday productivity, while critics warn that such personalization may blur the line between convenience and data exploitation. Read more

Google unveiled Gemma 4 12B, a 12‑billion‑parameter large language model that delivers near‑parity with its 26‑billion‑parameter sibling while fitting on a laptop with 16 GB of RAM. The model introduces Multi‑Token Prediction (MTP) drafters for faster, more efficient token generation and a streamlined vision embedding that removes bulky encoders. Audio inputs bypass encoding altogether, letting raw signals feed directly into the model. Weights, just under 18 GB, are downloadable from Kaggle and Hugging Face, and the model can be accessed through LM Studio, Google AI Edge Gallery and similar tools. Read more

Anthropic has named Morgan Stanley and Goldman Sachs as lead underwriters for its upcoming initial public offering, with JPMorgan Chase also on the syndicate. The company, which develops the Claude AI model, is eyeing an October listing after filing a confidential registration statement. A surprising detail emerged from SpaceX’s own IPO filing: the rocket firm provides Anthropic with roughly 325,000 Nvidia chips, a service that costs about $1.25 billion each month and runs through May 2029. The arrangement positions SpaceX as both supplier and competitor as the two firms prepare to go public. Read more

Alphabet completed an $85 billion equity offering, surpassing its original $40 billion target after demand flooded the market. The sale, which netted $45 billion in its first tranche, drew heavyweight investors including Berkshire Hathaway. CEO Sundar Pichai said the proceeds will back a multi‑year strategy to expand Google’s artificial‑intelligence infrastructure and data‑center capacity. The unprecedented raise signals strong institutional appetite for AI‑related assets and sets a new benchmark for public offerings in the sector. Read more

Google Cloud and Stockholm‑based startup Lovable announced an expanded multiyear collaboration on Wednesday that will increase Lovable's cloud usage fivefold, including a boost in AI workloads. The deal grants Lovable deeper access to Anthropic’s Claude and Google’s Gemini models, places its new agent in Google’s enterprise marketplace, and ties in security tools from Wiz, Google’s recent $32 billion acquisition. Lovable, which recently reported $400 million in annualized revenue and a $100 million month, counts more than half of the Fortune 500 among its users. The partnership aims to lock in enterprise customers while feeding Google’s massive cap‑ex plans for the year. Read more

Suno, the AI‑driven music‑generation startup, announced a $400 million Series D financing that pushes its valuation to $5.4 billion, just seven months after a $2.45 billion round. The funding comes as the company fights copyright suits from major labels that allege it trained its models on tens of thousands of protected songs. Suno maintains its approach is covered by fair‑use doctrine, and investors appear unfazed by the legal battles. Read more