AI regulation, antitrust, privacy law, and the rules shaping the tech industry.
61 articles
Manufacturers selling connected products in the EU must now alert ENISA and a national CSIRT within 24 hours of learning that a vulnerability is being actively exploited.
The revised law caps aggravated penalties at 10 percent of worldwide revenue, outranks GDPR limits, and writes the first national rules for training AI models on personal data.
California will build a public registry of AI auditors by 2029 and let independent verification organizations assess models under two bills endorsed by OpenAI and Anthropic.
Regulators on both sides of the Atlantic begin their reviews this week with one question at the center, whether the chip giant can control a neutral developer hub without abusing its power.
A mass takeover of a small wiki does not fit neatly into the law's categories of harm, and the Commission's decision to treat it as reportable will set the precedent for agent incidents.
Australia's proposal reaches beyond age limits to the recommendation engines themselves, daring platforms to defend opacity as a form of speech.
In a Statement of Interest the government calls model training transformative and warns that a loss could aid Chinese rivals, while The New York Times and other publishers ask a judge to reject that defense.
Backers point to recent loss-of-control incidents at major labs as the reason for the pause, and they want a new federal agency plus a civilian advisory board to spend two years writing rules before the biggest training runs could resume under license.
The ruling is the third straight failure of the US government's tech-breakup push, sparing Google its biggest antitrust exposure while imposing transparency rules on its ad auction.
The one-year moratorium also restricts companion chatbots and software with student-facing AI, while high school students keep limited access, as the district positions the pause as a study of how the tools affect child development.
The military's AI platform now spans three competing commercial providers, a deliberate move away from dependence on a single model vendor. The addition came with the highest security accreditation a non-classified system can receive, and the one major exclusion still tied up in litigation frames the launch as a policy story as much as a technical one.
The European Commission has placed OpenAI's ChatGPT into the Digital Services Act's strictest regulatory tier, a classification based on 159.1 million monthly users in the EU. The decision extends the bloc's internet rules to AI chat services and sets a precedent beyond Europe.
Meta agreed on August 26 to pay up to $16.7 billion over 10 annual installments to resolve claims from 29 states that it engineered Facebook and Instagram to addict young users, adding a midnight-to-6 a.m. lockout and a two-hour daily cap for teen accounts.
The EU AI Act's general-purpose AI obligations took effect on August 2, 2026, banning manipulative AI systems and imposing strict high-risk rules with fines up to 7% of global turnover, accelerating the Brussels Effect across global tech supply chains.
U.S. District Judge Rita F. Lin ruled the Defense Department's designation of Anthropic as a supply-chain risk was unlawful retaliation that violates the First Amendment, ordering all directives against the company rescinded.
Japan's government adopted guiding principles on intellectual property protection for generative AI operators on August 26, calling on AI businesses to publicly disclose training data types, collection methods, and copyright-infringement risk indicators.
Prime Minister Christopher Luxon's government introduced a bill on August 25 requiring age checks on Instagram, Snapchat, TikTok, and Facebook, with fines of up to 10% of global revenue for noncompliant platforms, but two coalition partners have said they will vote against it.
President Trump hosted crypto industry executives at the White House on August 19, 2026, calling on the Senate to pass the Clarity Act and reaffirming his administration's pro-crypto posture. The event marked the most direct presidential engagement with the digital asset industry since the 2025 Strategic Bitcoin Reserve executive order.
On August 2 the European Commission gained the power to fine GPAI providers up to €15 million or 3 percent of global turnover. The same week, FTC Chairman Andrew Ferguson said competition in AI models is healthy but monopoly risk in chips, power and construction is the real concern.
The Trump administration on August 18 sanctioned ICC President Tomoko Akane and Senior Trial Lawyer Abdoulaye Seye, the first time a sitting ICC president has been targeted by U.S. sanctions. Japan's government expressed regret; the ICC called the action a 'flagrant attack' on judicial independence.
The EU AI Act's prohibitions and general-purpose AI obligations came into general application on August 2. FASB on August 19 issued an exposure draft that would let companies classify compliant stablecoins like PYUSD as cash equivalents. The combined signal is that 2026 is the year frontier-AI and stablecoin compliance became operational rather than aspirational.