The Effros-Ruan conjecture for bilinear forms on \(C^{*}\)-algebras (Q944253)

This remarkable paper solves the main problem left open in the previous work by \textit{G.\,Pisier} and \textit{D.\,Shlyakhtenko} on the same subject [Invent.\ Math.\ 150, No.\,1, 185--217 (2002; Zbl 1033.46044)]. Namely, it proves the following Grothendieck inequality for bilinear forms on C*-algebras, in the framework of operator spaces. Let \(A,B\) be C*-algebras and let \(u:A\times B\to\mathbb C\) be a jointly completely bounded bilinear form. Then there exist states \(f_1,f_2\) on \(A\) and \(g_1,g_2\) on \(B\) such that \[ |u(a, b)|\leq \|u\|_{\text{jcb}}\left(f_1(aa^*)^{1/2}g_1(b^*b)^{1/2} +f_2(a^*a)^{1/2}g_2(bb^*)^{1/2}\right) \] for any \(a\in A\) and \(b \in B\), thereby confirming a conjecture of Effros-Ruan and Blecher in full generality. \textit{G.\,Pisier} and \textit{D.\,Shlyakhtenko} [loc.\,cit.]\ had proved the inequality above under the additional assumption that \(A\) or \(B\) is exact. The proof given here relies on the Grothendieck-Haagerup-Pisier inequality for C*-algebras in the Banach space setting and on special properties of Powers' III\(_{\lambda}\)-factors.