P ess Release
Nokia Global Netwo k T affic Repo t fo ecasts AI will d ive ea ly half of global t affic i 2035
- AI t affic behaves diffe e tly tha t aditio al applicatio t affic, eshapi g etwo k equi eme ts.
- Physical a d age tic AI will be amo g the fastest g owi g etwo k t affic catego ies i c easi g dema d fo upli k capacity, low late cy a d dist ibuted i f ast uctu e.
4 Novembe 2026
Espoo, Fi la d – Nokia today published its Global Netwo k T affic Repo t, fo ecasti g that AI will d ive ea ly half of global etwo k t affic i 2035. The fi di gs show etwo ks will i c easi gly eed to se ve co ected machi es as well as people, with AI d ivi g g eate upli k dema d, faste espo se times a d mo e p ocessi g close to use s a d devices.
By 2035, Nokia p ojects that total etwo k t affic will ea ly quad uple f om 2026 levels to almost 3,789 exabytes (EB) pe mo th, with 44% of that t affic d ive by AI. The i c easi g adoptio of auto omous AI age ts a d co ected machi es will challe ge etwo ks built p ima ily fo dow loads, i c easi g dema d fo upload capacity a d eal-time espo ses. Physical AI applicatio s, i cludi g obots a d vehicles that i te act with thei su ou di gs, a e p ojected to ge e ate mo e tha 90% of thei t affic ove the etwo k upli k by 2035.
Sta ti g f om a small base, age tic AI t affic is p ojected to each 160 times its 2026 levels i 2035, while physical AI t affic is expected to each 105 times its 2026 level ove the same pe iod. These p ojected wo kloads will equi e etwo ks that suppo t diffe i g eeds, f om bulk uploads to co ti uous, eal-time co ectio s.
The epo t also fo ecasts a shift i whe e AI data is p ocessed by 2035. While most AI t affic is expected to emai withi the ce t al cloud, edge a d met o co e i f ast uctu e is p ojected to se ve 36% of all AI- ative t affic by 2035 as age tic a d physical AI applicatio s ise i p omi e ce. Mo e efficie t models, low-late cy equi eme ts a d data sove eig ty eeds a e expected to b i g i tellige ce a d i fe e ci g close to use s a d devices, suppo ti g a mo e dist ibuted i f ast uctu e k ow as the AI g id.
Beyo d AI, the epo t p ojects that fixed wi eless access (FWA) will ge e ate mo e tha o e-thi d of co sume cellula t affic by 2035 as co ectio s a d usage g ow, expa di g cellula etwo ks’ ole i home b oadba d alo gside mobile se vices.
Nokia’s Global Netwo k T affic Repo t (2026–2035) uses a bottom-up model of co sume , e te p ise a d AI t affic. It combi es applicatio usage, co ected-device activity a d age tic wo kflows with thi d-pa ty esea ch, expe t eview a d Nokia lab testi g a d i sights to assess t affic g owth a d cha acte istics, etwo k dist ibutio a d pote tial upli k co st ai ts.
Pallavi Mahaja , Chief Tech ology a d AI Office at Nokia: “AI is cha gi g what we eed f om etwo ks, ot just how much t affic they ca y. Netwo ks built fo t aditio al i te et applicatio s will i c easi gly have to suppo t machi es that place ew dema ds o the upli k a d equi e eal-time espo ses. The ext phase of AI will depe d as much o the etwo k as o compute. Netwo k eadi ess must the efo e be a ce t al pa t of AI pla i g, ot a afte thought. At Nokia, we’ e buildi g the dist ibuted, AI- eady etwo ks that make this ext phase possible.”
Resou ces a d additio al i fo matio Webpage: Nokia Global Netwo k T affic Repo t (2026-2035)
About Nokia Nokia is a global leade i co ectivity i f ast uctu e, e abli g the AI supe cycle. We build etwo ks that co ect i tellige ce a d delive a comp ehe sive co ectivity solutio ac oss fixed, mobile, IP, optical, co e, a d data ce te etwo ks. With secu ity by desig , AI-d ive i ovatio , a d AI-optimized a chitectu e, Nokia delive s t usted i f ast uctu e that e ables AI to scale a d adva ces co ectivity to secu e a b ighte wo ld.
Media I qui ies Nokia P ess Office Email: P ess.Se vices@ okia.com
Follow us o social mediaLi kedI X I stag am Facebook YouTube
