Claude Fable 5 Hawadaa Laga Saaray !

Warbixin gaaban oo aan ku sheegayo in model-ka Fable 5 ee Anthropic la joojiyay, iyo sababta. Waxaan ugu talagalay qofka isticmaalaya models-ka Claude oo la yaabay markii access-kii ka go'ay.

Dawladda Maraykanka ayaa ku amartay Anthropic in access-ka Fable 5 laga joojiyo qof kasta oo aan Maraykanka u dhalan, xataa haddii uu halkaas ku nool yahay oo shirkadda u shaqeeyo. Sababtu waxay tahay in model-ka lagu helay jailbreak, taasoo ah qaab lagu dhaafo security guard-yada la hor dhigay, dawladduna waxay ku tilmaantay national security issue. Waxaan sharraxayaa in Anthropic ay horeba u qortay blog dheer oo ku saabsan in dawladdu awood u yeelato inay models joojiso, iyo in CEO-gooda uu had iyo jeer riixayay in dawladdu policies u dejiso. Waxaan sidoo kale sheegayaa in shirkadaha kale, GPT-5.5 oo ay ku jirto, ay isla ciladdaas leeyihiin, iyo in researchers Amazon ka tirsan ay arrinta soo warramtay.

Ugu dambayn waxaan kaga hadlayaa waxa arrintani u dhigan tahay mustaqbalka: haddii model walba uu marayo checking dawladeed, xawaaraha AI research-ka wuu hoos u dhici karaa. Waxaan sidoo kale sharraxayaa sababta Yurub ay uga daahday, iyo in Opus uu meesha ka buuxinayo.

A 5-minute Somali-language (Af-Soomaali) news video from Dugsiiye on Anthropic's Fable 5 being pulled. It reports that the US government ordered access cut off for anyone not born in the United States, including staff living there, after a jailbreak was found in the model and the issue was treated as national security. The video sets this against Anthropic's own published position that governments should be able to halt models, notes that the same jailbreak class also affects other vendors' models including GPT-5.5, and that Amazon researchers reported the finding. It closes on what the precedent means: if every new model has to clear a government review, the pace of AI research slows, and it contrasts this with Europe's policy-first approach and its single French model. Free on YouTube.