most all the big AI news this year was about how fast the technology is progressing , the harm it ’s causing , and surmise about how shortly it will develop past the point where humans can control it . But 2024 also saw government make significant inroads into govern algorithmic systems . Here is a breakdown of the most crucial AI legislation and regulatory efforts from the past twelvemonth at the state , Union , and international levels .
State
U.S. United States Department of State lawmakers took the lead on AI regulation in 2024 , introducinghundreds of bills — some had modest goals like creating study committees , while others would have imposed serious polite financial obligation on AI developer in the event their creations cause catastrophic harm to society . The vast majority of the billhook failed to go on , but several body politic enacted meaningful legislation that could serve as mannequin for other states or Congress ( assuming Congress ever starts functioning again ) .
Unsurprisingly , given that it ’s the backyard of the tech industry , some of the most challenging AI proposal came out of California . One high - profile bill would have impel AI developers to take base hit forethought and held companies liable for catastrophic indemnity due to their system . That visor pass along both bodies of the legislature amid a fierce lobbying effort but wasultimately vetoedby Governor Gavin Newsom .
Newsom did , however , signal more than a dozenother billsaimed at less apocalyptic but more immediate AI harms . One newfangled California law requires wellness insurance firm to ensure that the the AI systems they apply to make insurance coverage determinations are fair and equitable . Another requires generative AI developer to produce tool that recording label contentedness as AI - generated . And a twosome of bank bill prohibit the statistical distribution of a dead person ’s AI - generated semblance without anterior consent and mandate that concord for support peoples ’ AI - yield likenesses must clearly specify how the content will be used .
The AI executive order Joe Biden signed in 2023 paved the way for much of the federal government’s work in 2024.© Bloomberg via Getty Images
Colorado passed afirst - of - its - kind in the U.S. lawrequiring companies that produce and use AI systems to take sane footprint to assure the tool are n’t discriminatory . Consumer advocates call the legislating animportant service line . It ’s likely that similar bills will be hotly debated in other state in 2025 .
And , in a in-between fingerbreadth to both our future robot overlord and the planet , Utahenacted a lawthat proscribe any governmental entity from granting legal personhood to artificial intelligence , pulseless target , bodies of water , atmospherical gases , weather , plant life , and other non - human things .
Federal
Congress talked a lot about AI in 2024 , and the House ended the year by releasing a273 - pageboy two-way reportoutlining conduct principles and recommendations for future regulation . But when it came to really pop off legislation , federal lawgiver did very little .
Federal agency , on the other hand , werebusy all yeartrying to get in touch with the goals set out in President Joe Biden ’s 2023 executive director rescript on AI . And several regulator , particularly the Federal Trade Commission and Department of Justice , crack down on deceptive and harmful AI system .
The oeuvre government agency did to comply with the AI executive order was n’t especially aphrodisiacal or headline grabbing , but it laid important foundations for the governance of public and secret AI systems in the future . For example , federal agencies ship on an AI - talent lease spree and createdstandardsfor creditworthy example development and trauma mitigation .
And , in a self-aggrandising step toward increasing the public ’s sympathy of how the political science uses AI , the Office of Management and Budget wrangled ( most of ) its fellow agencies into disclosingcritical informationabout the AI organisation they use that may bear on hoi polloi ’s rights and safety .
On the enforcement side , the FTC’sOperation AI Complytargeted companies using AI in deceptive ways , such as to write fake reviews or furnish effectual advice , and itsanctionedAI - gun detection company Evolv for making misleading claim about what its merchandise could do . The agency alsosettledan investigation with facial recognition fellowship IntelliVision , which it accuse of falsely allege its engineering was liberal of racial and gender prejudice , andbannedthe pharmacy chain Rite Aid from using facial realization for five years after an probe regulate the society was using the peter to discriminate against shoppers .
The DOJ , meanwhile , joined state lawyer general in a lawsuit accusing the real estate software companyRealPage of a massive algorithmic terms - fixing schemethat raised rent across the nation . It also won several anti - trust lawsuits against Google , include one involving the company’smonopoly over net searchesthat could importantly shift the balance of power in the burgeon AI hunt diligence .
Global
In August , the European Union ’s AI Actwent into effect . The law , which is already serving as a model for other jurisdictions , requires AI arrangement that do high-pitched - peril map , such as assisting with hiring or aesculapian decisions , to undergo danger mitigation and meet sure standards around breeding data caliber and human oversight . It also bans the use of other AI systems , such as algorithms that could be used to attribute a country ’s occupier societal scores that are then used to deny right and privileges .
In September , China bring out a major AI safety governanceframework . Like similar framework write by the U.S. National Institute of Standards and Technology , it ’s non - binding but creates a vulgar band of standard for AI developers to survey when identifying and palliate risks in their systems .
One of the most interesting pieces of AI policylegislationcomes from Brazil . In late 2024 , the body politic ’s senate passed a comprehensive AI safety bill . It faces a challenging road forwards , but if passed , it would create an unprecedented set of protections for the variety of copyright fabric commonly used to train generative AI system . Developers would have to disclose which copyrighted stuff was included in their training data , and creators would have the major power to disallow the manipulation of their study for training AI systems or negotiate compensation agreements that would be based , in part , on the size of the AI developer and how the material would be used .
Like the EU ’s AI Act , the proposed Brazilian natural law would also require high - jeopardy AI systems to follow certain safety communications protocol .
AItech policy
Daily Newsletter
Get the best technical school , science , and culture tidings in your inbox day by day .
News from the future , delivered to your present tense .
You May Also Like
