Regulation & Standards
1 gain3 Traces39 problems
Every measured gain and problem on the record for this topic, newest first.Too few to weigh yet. Updated .
AI regulation and standards
43 records so far, too few to weigh yet. Is your study missing? Add it
A Trace is one record with both a gain and a problem, so it counts on both sides: 4 records show a gain and 42 show a problem. 59 context records are in the record list but not counted, because they report no measured gain or problem.
| Evidence | Gain side | Problem side |
|---|---|---|
| HighStrong study design, backed by at least two independent sources | 0 | 1 |
| ModerateA measured study or an official source | 4 | 19 |
| EmergingEarly, small, unreviewed or reported second-hand | 0 | 22 |
By type of evidence: 10 investigations by reporters; 9 news reports; 9 commentaries or narrative reviews; 8 regulatory or legal actions; 4 interview studies; 2 systematic reviews or meta-analyses; 1 observational study. How evidence is scored
43 records so far, too few to weigh yet. A written summary appears once a topic has at least 20 counted records, with at least 5 on each side. Is your study missing? Add it
A review of 43 studies (2020-2025) found AI improved efficiency and service delivery in government decisions, through automation of routine tasks and predictive analytics.
FDA says placing AI-assisted Pap test slide review systems in class II instead of automatic class III should widen patient access to the devices by lowering regulatory burdens.
California's new package, which Gavin Newsom calls nation-leading, sets what supporters describe as the strongest U.S. companion chatbot rules and bans addictive features like infinite scroll for under-16s to protect young users from harm.
Germany's regulator says AI summaries sit prominently in results, cut the visibility of traditional links and hurt sales of third-party media content, while Google faces liability for inaccurate AI-generated claims.
A review of six studies found most heart-care AI research looked at clinical or economic performance without referencing formal health technology assessment frameworks, and cardiologist engagement was inconsistently reported.
The European Commission preliminarily found that TikTok's recommendation algorithm keeps serving repetitive content on themes like body image and isolation, and that its safeguards, including the "not interested" button, failed to stop it for the platform's 189 million EU users.
102 records · page 4 of 7
On Tuesday, September 30, 2026, President Trump and six AI and tech executives signed an accord on AI safety. The copy released by the White House misspelled the president's title as "President of the Unites States", and it is unclear how that happened or whether the error will reach the Federal Register.
The Guardian
In an interview published in October 2026, AI pioneer Geoffrey Hinton repeated his warnings that advanced AI systems could escape human control.
California's governor signed a package of laws restricting how employers use AI. They bar firing decisions made entirely by AI, emotion prediction, neural data collection and AI bathroom surveillance, and require notice…
On October 4, 2026, President Trump named Jay Clayton, who already directs national intelligence, as the White House AI czar. Clayton will lead a body Trump calls the Super Intelligence Force, which is meant to…
In an interview for Politico's Decoded podcast released on Monday 5 October 2026, OpenAI's chief executive argued that a lighter regulatory approach means accepting that some harms will occur.
George Williamson, the new head of the Alan Turing Institute, says the UK should build its own AI capability instead of relying on foreign systems.
The article is a bioethical analysis of how federal and state regulation in the United States handles AI in the digital spaces where adolescents meet health information, including chatbots, AI-enabled…
Microsoft president Brad Smith told the BBC that AI is governed by weaker safety rules than washing machines or milk.
Notes seen by Guardian Australia show a senior adviser to Anthony Albanese edited and apparently approved wording in a bipartisan parliamentary committee letter on AI policy before it was sent.
Diagnos Inc recalled its CARA (Computer Assisted Retinal Analysis) software, version 4.7.2, after identifying that some features built on artificial intelligence potentially lack appropriate market…
Researchers screened official international and European Union documents in April and May 2025 to examine how ethics and regulation govern the behavior of AI systems that analyze medical images.
The authors describe a two-day Utah forum in September where randomly selected residents deliberated on the public's role in decisions about AI and datacentres.
A Reform UK member of the Senedd gave a five-minute, statistic-heavy speech on 20mph zones without glancing at notes, while flashes in his spectacles led to suspicion that smart glasses were acting as…
Published September 2023 in Public Policy and Administration, the article presents a conceptual framework for a third wave of digital era governance driven by data science and AI.
On 5 October 2026 The Guardian published a comment piece arguing that Andy Burnham's government must decide priorities for the UK's G20 presidency next year.
ace