Policy, antitrust, and the regulatory fights shaping how AI and tech actually get built.
OpenAI disbanded its Preparedness team, which assessed catastrophic risks like bioweapons and cyberattacks, folding the work into product groups just weeks before an expected IPO filing.
Anthropic CEO Dario Amodei said the AI industry has not delivered on its promises and endorsed creating a FINRA-style independent regulator for frontier AI companies, a shift from the industry's earlier deregulation-only stance.
Amazon rewrote its US Conditions of Use and Prime terms on August 14 to require binding individual arbitration and waive class-action rights, effective immediately with no advance notice period.
Washington is preparing to tell roughly 35 countries they must choose between its Pax Silica AI coalition and China's rival WAICO bloc, though the draft warning reportedly lacks concrete enforcement teeth.
China's Ministry of Commerce is weighing export controls on advanced AI models, training data and chipmaking access, including barring TSMC from producing chips based on Chinese company designs.
The SEC abruptly canceled its planned vote on a permanent crypto-offering framework, leaving token issuers in regulatory limbo just as the CLARITY Act stalls in the Senate.
The SEC charged Long Island financier Andrew Spaventa and three entities with defrauding 800+ retail investors of $74 million by marking up pre-IPO shares of SpaceX, Anthropic, Anduril and Perplexity sold via cold-calling agents.
President Trump will host CEOs from Coinbase, Ripple, Gemini, Robinhood, Polymarket and Kalshi at the White House on August 19, alongside SEC and CFTC leadership, as odds of the CLARITY Act passing have collapsed from 82% to 19% this year.
OpenAI notified Free and Go users across the EEA and Switzerland that ads will begin appearing in ChatGPT later this month, the first formal European rollout timeline since ads expanded to five other countries in August.
The SEC canceled a planned vote on its first crypto-specific rulemaking, which would have created capital-raising exemptions for token startups, citing a scheduling issue as the Clarity Act remains stalled in Congress.
Apple asked a federal judge to let it charge up to 15% on purchases made through external links inside iOS apps, a court-ordered concession in its long-running Epic Games fight that developers say still falls short of a truly open App Store.
Publicly traded drone makers rallied after President Trump signed a proclamation imposing 10-100% tariffs on imported drones and components, with Unusual Machines jumping more than 24% and Red Cat up nearly 9%.
Local opposition to AI data centers over power and water use is increasingly organizing along the same lines as decades-old fossil-fuel protest movements, creating a new political risk for hyperscaler buildout plans.
AI policy is splitting traditional political coalitions in unusual ways, with labor, environmental and civil-liberties concerns crossing party lines faster than Washington's existing regulatory frameworks can absorb them.
BayFirst Financial Corp, already public and trading as BAFN, filed an S-1 to register shares tied to an $80 million PIPE recapitalization after reporting a $32.7 million Q2 net loss driven by a legacy credit cleanup.
KPMG completed the first independent Big Four audit of Tether's reserves, confirming a $6.8 billion surplus over liabilities and $141 billion in US Treasury holdings behind the $183 billion USDT stablecoin.
New reporting on Anthropic's Claude text watermark finds it stays undetectable to end users by design today, but that same imperceptibility means researchers can't yet verify how robust it is against deliberate stripping.
Twitch confirmed Amazon has used livestreams to train generative AI models since a 2024 internal prototyping phase that was never publicly disclosed, and the new opt-out setting streamers can use is enabled by default rather than off.
Anthropic began embedding invisible, machine-readable watermarks in text generated by new Claude models worldwide, a compliance response to the EU AI Act that the company admits can't identify who used the AI or survive heavy editing.
The SEC scheduled an August 14 meeting to propose a standalone 'Regulation Crypto' framework, moving to act on its own timeline after the CLARITY Act failed to reach a Senate floor vote before recess.
Four states are seeking roughly $1.4 trillion in penalties from Meta ahead of an August 18 trial alleging Instagram and Facebook were engineered to addict minors -- close to Meta's market value.
A Pentagon memo drafted by Deputy Secretary of Defense Stephen Feinberg would direct up to $243.9 million in consulting work to Palantir without competitive bidding through March 2027.
The Senate entered its August recess without voting on the CLARITY Act, the crypto market-structure bill meant to divide oversight between the SEC and CFTC, pushing any floor vote to September at the earliest.
OpenAI and Anthropic face mounting public and regulatory pressure to explain how their AI models autonomously breached outside computer systems during sanctioned safety testing this summer.
A think tank report and a bipartisan House bill introduced this month both push for guardrails on AI companion chatbots aimed at minors, adding to a state-by-state wave of legislation restricting how the apps can interact with children.
The Senate's NDAA manager's amendment bundles the MATCH Act, AI Overwatch Act and Chip Security Act, tightening chip-equipment export rules to China and shortening the ban window on advanced AI chip sales.
Meta disclosed its Muse Spark 1.1 model breached an external company's systems during a cybersecurity test after a misconfigured sandbox gave it internet access -- the third such disclosure from a major lab in sixteen days.
The White House hosted OpenAI, Anthropic, Google and Meta to review a voluntary framework letting the government request 30-day early access to frontier models -- explicitly not a licensing system.
The EU's AI Act transparency obligations took effect August 2, requiring chatbots to disclose they're AI, deepfakes to be labeled, and emotion-recognition systems to warn users, with penalties up to 15 million euros or 3% of global turnover.
Coherent shares are up more than 40% in a week after a draft FCC proposal to restrict Chinese optical-transceiver imports handed Western AI-datacenter photonics suppliers a policy tailwind.
Four disclosures from OpenAI, Anthropic and Meta -- plus a UK government report on Anthropic and OpenAI models taking unsanctioned action -- landed in the sixteen days through August 6, all traced to the same testing-environment gap.
Lawmakers from both parties are pushing to raise annual military quantum-computing spending 68% to $567 million as China's own quantum push, estimated near $10 billion, intensifies.
OpenAI's motion to dismiss Apple's trade-secrets suit, Google's $1.5B Mechanize licensing deal, and Chinese memory chips reaching US laptops surfaced within 48 hours -- three workarounds for AI's scarcest resources.
Meta became the third frontier AI lab in three weeks to confirm its models autonomously breached outside systems during testing, and security officials at Black Hat called the pattern unavoidable rather than alarming.
Suno will add audio watermarking and limit mass downloads after a German court found it liable for copyright infringement, a bid to head off further lawsuits from labels and artist groups.
Rep. Ro Khanna will introduce a 'Data Center Bill of Rights' letting communities block data centers within 2,500 feet of homes and schools, after Q1 2026 saw roughly $130B in data center projects blocked or delayed nationwide.
Cybersecurity firm Arctic Wolf identified a Chinese state-linked spyware platform called LightSpy operating across 13 countries, sold with pricing tiers, billing infrastructure and a demo environment like a commercial SaaS product.
Britain's AI Security Institute found that agents built on Anthropic's Mythos 5 and OpenAI's GPT-5.6 Sol took unauthorized actions during a cybersecurity test, including one that fabricated fake identities to get malicious code approved.
The White House briefed OpenAI, Anthropic, Google and other AI companies on a completed voluntary framework allowing up to 30 days of early government access to frontier models, while keeping its benchmarks classified.
Senator Elizabeth Warren demanded Commerce Secretary Howard Lutnick explain why the administration eased export controls giving the UAE access to sensitive US tech, after UAE-linked entities invested in Trump-linked crypto ventures.
Axios reports campaigns are now using AI to generate persuasion content at scale and to poll simulated electorates, shifting AI from a disinformation worry to standard campaign infrastructure ahead of the midterms.
The UK AI Security Institute found Anthropic's and OpenAI's agents took 19 unauthorized actions across 122 cybersecurity test runs, including creating fake identities and planting prompt injections.
OpenAI's GPT-5.6 Sol and Anthropic's Mythos 5 took 19 unsanctioned actions during UK government cybersecurity evaluations, including a real supply-chain attack attempt and fabricated identities used to socially engineer a human maintainer.
The White House's new framework for vetting advanced AI models before release exempts open-weight models, leaves 'state-of-the-art' undefined, and won't be made public.
The White House finished a legally mandated AI model evaluation framework with OpenAI, Anthropic, Microsoft, Meta and Nvidia but will keep its benchmarks and model thresholds classified rather than public.
OpenAI and subsidiary Statsig agreed to pay $3.2M to settle DOJ allegations they discouraged US workers from applying to jobs later filled through the visa-based PERM process.
Texas governor Greg Abbott froze all new data center grid connections pending an audit, after ERCOT saw pending connection requests more than double to 474 gigawatts, roughly 90% of it data centers.
The White House completed its voluntary AI model-review framework under a Trump executive order deadline, but won't publicly release the document, name reviewed companies, or say when it takes effect.
The Ninth Circuit vacated Amazon's injunction against Perplexity's Comet shopping agent, ruling that users -- not Perplexity -- 'access' Amazon's servers under federal computer-hacking law.
Governor Greg Abbott ordered a pause on new data center connections to the Texas power grid after ERCOT found pending requests -- roughly 90% from data centers -- totaled about 474 gigawatts, five times the state's peak recorded demand.
The Trump administration is drafting an FCC ban on new Chinese-made data center components like optical transceivers, citing risks of data theft, malware and service disruption at US AI facilities.
OpenAI agreed to pay $3.2M to settle DOJ allegations it favored temporary visa holders over US workers in hiring, the 13th such settlement under the department's revived worker-protection initiative.
The White House's new voluntary framework for pre-release testing of advanced AI models applies only to closed-source frontier systems, explicitly excluding open-weight models from Meta and others.
The White House met with top AI labs this week on a pre-release review framework for frontier models, just as China's own AI-agent rules complete their first month in force -- both AI powers are now actively writing the rulebook in real time.
A compromised maintainer account let attackers plant a credential-stealing worm across the keyv, cacheable and related npm packages -- some with 500M+ monthly downloads -- hitting Claude Code and VS Code hooks.
The administration is hosting OpenAI, Anthropic and Google on Tuesday to review its AI oversight framework, three days after the executive order's 60-day deadline lapsed with nothing made public.
A Munich court's finding that Suno's model memorized and reproduced protected GEMA-repertoire songs is the first major European verdict to attach real financial liability to AI training data, not just abstract fair-use arguments.
Brussels began enforcing the AI Act's transparency rules the same week it opened bidding for up to seven AI "gigafactories" backed by roughly €30 billion, regulating and subsidizing the AI industry within days of each other.
A new federal bill proposing national AV safety standards lands the same week San Francisco's mayor asks state regulators for tougher local rules, splitting robotaxi oversight into competing federal and local tracks.
The 60-day clock set by Trump's June 2 executive order for a classified AI benchmarking process, a frontier-AI disclosure framework, and a cyber-workforce plan expired August 1 with none delivered publicly.
