Pope calls for robust regulation of AI in manifesto that ponders the future of humanity

Pope Leo XIV called Mon­day for ro­bust reg­u­la­tion of ar­ti­fi­cial in­tel­li­gence and for its de­vel­op­ers to work for the com­mon good rather than prof­it, is­su­ing a sweep­ing man­i­festo on safe­guard­ing hu­mankind as the tech­nol­o­gy im­pacts every­thing from work to war.

“Mag­nifi­ca Hu­man­i­tas” (Mag­nif­i­cent Hu­man­i­ty), Leo’s first en­cycli­cal, has been ea­ger­ly await­ed ever since his­to­ry’s first U.S.-born pope an­nounced days af­ter his elec­tion that he con­sid­ered AI to be the biggest chal­lenge fac­ing hu­man­i­ty to­day.

In the text, Leo de­nounced the “cul­ture of pow­er” dri­ving the AI race, es­pe­cial­ly in de­vel­op­ing ever more so­phis­ti­cat­ed meth­ods of re­mote war­fare. He de­clared that it was “not per­mis­si­ble” to en­trust ir­re­versible, lethal de­ci­sions to AI sys­tems, set­ting up an­oth­er flash point be­tween the Amer­i­can pope and the Trump ad­min­is­tra­tion, which has worked ag­gres­sive­ly to dereg­u­late AI de­vel­op­ment.

“Ar­ti­fi­cial In­tel­li­gence now de­mands to be dis­armed, freed from log­ics that turn it in­to an in­stru­ment of dom­i­na­tion, ex­clu­sion and death,″ the pope told a spe­cial Vat­i­can pre­sen­ta­tion of the en­cycli­cal, one of the most au­thor­i­ta­tive types of teach­ing doc­u­ments a pope can is­sue.

Ex­perts in the tech in­dus­try, acad­e­mia and Catholic moral­i­ty said the doc­u­ment will like­ly be­come a bench­mark in the de­bate over AI, a point of ref­er­ence for pol­i­cy­mak­ers, re­searchers and or­di­nary folk alike. It comes as the near-dai­ly de­vel­op­ments in the tech­nol­o­gy trig­ger con­cerns over AI re­plac­ing hu­man jobs and even hu­man in­tel­li­gence.

Tay­lor Black, a Mi­crosoft AI ex­ec­u­tive and di­rec­tor of Catholic Uni­ver­si­ty of Amer­i­ca’s AI in­sti­tute, said the doc­u­ment would prompt peo­ple “at the fore­front of these tools” to ask ques­tions such as “What does it mean to be hu­man?”

Pope calls out AI com­pa­nies even as he hosts An­throp­ic

The Vat­i­can launch al­so in­clud­ed re­marks by the co-founder of An­throp­ic, which is cur­rent­ly locked in a le­gal bat­tle with the Trump ad­min­is­tra­tion over ac­cess to its AI tech­nol­o­gy. The Vat­i­can de­cid­ed to in­volve An­throp­ic as part of its decade-long ef­fort to en­gage Sil­i­con Val­ley in di­a­logue over the hu­man cost of AI.

And yet in his text, Leo re­peat­ed­ly blast­ed the con­cen­tra­tion of pow­er and da­ta in the hands of so few peo­ple in the pri­vate sec­tor as a dan­ger, es­pe­cial­ly to chil­dren and the most vul­ner­a­ble, and called for ex­ter­nal reg­u­la­tion of their work.

“It is not enough to in­voke ethics in the ab­stract; ro­bust le­gal frame­works, in­de­pen­dent over­sight, in­formed users and a po­lit­i­cal sys­tem that does not ab­di­cate its re­spon­si­bil­i­ty are re­quired,” he wrote. “A more moral AI is not enough if that moral­i­ty is de­ter­mined by a few.”

Leo ap­pealed to AI de­vel­op­ers and po­lit­i­cal lead­ers re­spon­si­ble for reg­u­lat­ing them to slow down and re­flect on what they are do­ing. He urged them to use eth­i­cal and spir­i­tu­al guide­lines to make the choice to work not for their own prof­it or pow­er, but the bet­ter­ment of hu­man­i­ty.

AI com­peti­tors Ope­nAI and An­throp­ic are the sec­ond- and third-most valu­able U.S. pri­vate com­pa­nies, each val­ued at hun­dreds of bil­lions of dol­lars, more than the GDP of many na­tions. Both com­pa­nies are head­ing to­ward near-tril­lion-dol­lar IPOs.

An­throp­ic co-founder Christo­pher Olah wel­comed Leo’s crit­i­cism and con­cern. He said such ex­ter­nal checks were fun­da­men­tal to the tech­nol­o­gy “go­ing well” for hu­mankind since there is so much at stake — “a re­al pos­si­bil­i­ty that AI will dis­place hu­man la­bor at a very large scale.”

“We need more of the world — re­li­gious com­mu­ni­ties, civ­il so­ci­ety, schol­ars, gov­ern­ments — to do what His Ho­li­ness has done here: to take this se­ri­ous­ly, to look close­ly, and to push events in a bet­ter di­rec­tion,” Olah said. “We need moral voic­es that the in­cen­tives can­not bend.”

Ex­perts say the text will be­come a bench­mark

In a me­thod­i­cal text, the math ma­jor pope traced the his­to­ry of the Catholic Church’s so­cial teach­ing and ap­plied its core con­cepts — jus­tice, sol­i­dar­i­ty, the dig­ni­ty of work and the uni­ver­sal des­ti­na­tion of re­sources — to the dig­i­tal rev­o­lu­tion.

“I am con­vinced that this will prove to be a defin­ing doc­u­ment for our era, a pro­found and prophet­ic doc­u­ment,” said Pao­lo Caroz­za, law pro­fes­sor at Notre Dame Law School and chair of the Meta Over­sight Board.

“Pope Leo is of­fer­ing a clear, com­pre­hen­sive, and co­her­ent voice urg­ing us to take re­spon­si­bil­i­ty for con­struct­ing a world in which tech­nol­o­gy will serve hu­mans rather than de­grade them,” he said.

In its strongest chap­ters, Leo de­nounced how AI had helped ac­cel­er­ate the “nor­mal­iza­tion of war” by de­sen­si­tiz­ing peo­ple to its cost. He didn’t name spe­cif­ic con­flicts, but cit­ed “op­pos­ing im­pe­ri­alisms, be­tween pow­ers that wish to pre­serve their su­prema­cy, and those that as­pire to seize that su­prema­cy.”

He de­mand­ed trans­paren­cy and ac­count­abil­i­ty by AI de­vel­op­ers so that the chain of de­ci­sion-mak­ing com­mand in or­der­ing strikes with AI weapon­ry is al­ways known. He de­clared that the Catholic Church’s “just war” the­o­ry, which pro­vides spe­cif­ic cri­te­ria for when force can be jus­ti­fied, was now “out­dat­ed” giv­en the tech­no­log­i­cal ad­vances of war­fare.

A text in the church’s so­cial jus­tice tra­di­tion

Leo signed the text May 15, the 135th an­niver­sary of the pub­li­ca­tion of “Re­rum No­varum” (Of New Things), the most im­por­tant teach­ing doc­u­ment of Leo’s hero and name­sake, Pope Leo XI­II. That doc­u­ment ad­dressed work­ers’ rights, the lim­its of cap­i­tal­ism, and the oblig­a­tions that states and em­ploy­ers owed work­ers as the In­dus­tri­al Rev­o­lu­tion was un­der­way.

It be­came the foun­da­tion of mod­ern Catholic so­cial thought, and the cur­rent pope cit­ed it at the start of his pon­tif­i­cate in re­la­tion to the AI rev­o­lu­tion, which he be­lieves pos­es the same ex­is­ten­tial ques­tions that the In­dus­tri­al Rev­o­lu­tion posed over a cen­tu­ry ago. “Mag­nifi­ca Hu­man­i­tas” thus be­comes the lat­est chap­ter in a cen­tu­ry-long his­to­ry of popes adapt­ing “Re­rum No­varum” to the so­cial ques­tions of their times, of­ten dwelling on the dig­ni­ty of work for hu­man flour­ish­ing.

AI is evok­ing both ex­is­ten­tial fears and utopi­an vi­sion amid an in­ten­si­fy­ing de­bate on whether it will be­come a cat­a­lyst that en­rich­es hu­man­i­ty or a tech­no­log­i­cal tox­in that dulls hu­man in­tel­li­gence while wip­ing out mil­lions of high-pay­ing jobs.

“The pur­suit of greater prof­its can­not jus­ti­fy choic­es that sys­tem­at­i­cal­ly sac­ri­fice jobs, be­cause the hu­man per­son is an end, not a means, and the eco­nom­ic or­der must re­main sub­or­di­nate to hu­man dig­ni­ty and the com­mon good,” Leo wrote.

Leo ex­tend­ed his con­cern for up­hold­ing hu­man dig­ni­ty in la­bor to is­sue the first-ever pa­pal apol­o­gy for the Holy See’s own role in le­git­imiz­ing slav­ery by giv­ing Eu­ro­pean sov­er­eigns ex­plic­it au­thor­i­ty to sub­ju­gate and en­slave “in­fi­dels.”

A decade-long di­a­logue with Sil­i­con Val­ley

Vat­i­can of­fi­cials de­clined to say who con­tributed to Leo’s en­cycli­cal. But Vat­i­can and church of­fi­cials have been en­gaged in a di­a­logue with Sil­i­con Val­ley tech firms for a decade.

The de­ci­sion to in­clude An­throp­ic at the Vat­i­can launch was crit­i­cized by some who con­sid­ered it a pa­pal stamp of ap­proval of the AI firm, which is cur­rent­ly su­ing the Trump ad­min­is­tra­tion af­ter it or­dered all U.S. agen­cies to stop us­ing An­throp­ic’s tech­nol­o­gy for its re­fusal to al­low the U.S. mil­i­tary un­re­strict­ed use of it.

Bri­an Boyd, U.S. faith li­ai­son for the non­prof­it Fu­ture of Life In­sti­tute, read the in­clu­sion of An­throp­ic’s co-founder Olah as a recog­ni­tion of its promi­nence in the field and as sim­i­lar to a pa­pal au­di­ence with a head of state: not an en­dorse­ment.

An­throp­ic is an “enor­mous cor­po­ra­tion that is tak­ing on­to it­self an enor­mous risk and re­spon­si­bil­i­ty,” Boyd said, adding that the com­pa­ny has “demon­strat­ed gen­uine good­will and in­tegri­ty and in­ter­est in di­a­logue.” —VAT­I­CAN CITY (AP)

__________

Sto­ry by NICOLE WIN­FIELD, KAIT­LYN HUA­MANI and PAO­LO SAN­TALU­CIA | As­so­ci­at­ed Press

Nicole Win­field re­port­ed from Mid­dle­town, Con­necti­cut, and Kait­lyn Hua­mani re­port­ed from Los An­ge­les. As­so­ci­at­ed Press writ­ers Kelvin Chan in Lon­don and Colleen Bar­ry in Mi­lan con­tributed to this re­port.