diff --git a/NEWS.md b/NEWS.md
index 5f825c023..17d97ee17 100644
--- a/NEWS.md
+++ b/NEWS.md
@@ -4,12 +4,14 @@
 
 ## data.table [v1.18.99](https://github.com/Rdatatable/data.table/milestone/37?closed=1)  (in development)
 
-### BREAKING CHANGE
+### BREAKING CHANGES
 
 1. `hasNA` argument to `froll*` functions no longer works. Use `has.nf` instead; see the release note for 1.18.0, item 3 under `BREAKING CHANGE`.
 
 2. `x` and `n` arguments to `frollapply` functions no longer work. Use `X` and `N` instead, respectively; see the release note for 1.18.0, item 4 under `BREAKING CHANGE`.
 
+3. options `"datatable.old.matrix.autoname"` is now `FALSE` by default, meaning `names(data.table(x=1, cbind(1)))` is now `c("x", "V2")`. Toggle the option to retain the old behavior for now; future releases will work to remove this possibility. See the release notes for 1.18.0, item 1 under `NOTE OF INTENDED FUTURE POTENTIAL BREAKING CHANGES`.
+
 ### Notes
 
 1. {data.table} now depends on R 3.5.0 (2018).
diff --git a/R/onLoad.R b/R/onLoad.R
index 12cc93db1..b72fee4d1 100644
--- a/R/onLoad.R
+++ b/R/onLoad.R
@@ -98,7 +98,7 @@
     datatable.auto.index=TRUE,          # DT[col=="val"] to auto add index so 2nd time faster
     datatable.use.index=TRUE,           # global switch to address #1422
     datatable.prettyprint.char=NULL,    # FR #1091
-    datatable.old.matrix.autoname=TRUE  # #7145: how data.table(x=1, matrix(1)) is auto-named set to change
+    datatable.old.matrix.autoname=FALSE # #7145: how data.table(x=1, matrix(1)) is auto-named set to change
   )
   opts = opts[!names(opts) %chin% names(options())]
   options(opts)
diff --git a/inst/tests/tests.Rraw b/inst/tests/tests.Rraw
index 8b23e099a..35727c3b5 100644
--- a/inst/tests/tests.Rraw
+++ b/inst/tests/tests.Rraw
@@ -21329,15 +21329,10 @@ if (test_R.utils) local({
 })
 
 # Create a data.table when one vector is transposed doesn't respect the name defined by user #4124
-local({
-  old = options(datatable.old.matrix.autoname=FALSE)
-  on.exit(options(old))
-
-  test(2321.01, DT <- data.table(a=1:2, b=matrix(1:2)), data.table(a=1:2, b=1:2))
-  test(2321.02, names(DT), names(data.frame(a=1:2, b=matrix(1:2))))
-  test(2321.03, DT <- data.table(a=integer(), b=matrix(1L, nrow=0L, ncol=1L)), data.table(a=integer(), b=integer()))
-  test(2321.04, names(DT), names(data.frame(a=integer(), b=matrix(1L, nrow=0L, ncol=1L))))
-})
+test(2321.01, DT <- data.table(a=1:2, b=matrix(1:2)), data.table(a=1:2, b=1:2))
+test(2321.02, names(DT), names(data.frame(a=1:2, b=matrix(1:2))))
+test(2321.03, DT <- data.table(a=integer(), b=matrix(1L, nrow=0L, ncol=1L)), data.table(a=integer(), b=integer()))
+test(2321.04, names(DT), names(data.frame(a=integer(), b=matrix(1L, nrow=0L, ncol=1L))))
 ## but respect named column vectors
 test(2321.05, DT <- data.table(a=1:2, cbind(b=3:4)), data.table(a=1:2, b=3:4))
 test(2321.06, names(DT), names(data.frame(a=1:2, cbind(b=3:4))))
@@ -21376,28 +21371,25 @@ test(2321.26, as.data.table(M), data.table(A=1:3, V2=4:6))
 test(2321.27, as.data.table(M, keep.rownames='id'), data.table(id=c('a', 'b', 'c'), A=1:3, V2=4:6))
 
 # also respect old auto-naming rules by default (to be deprecated)
-test(2321.28, names(data.table(a=1, cbind(2), c=3, 4)), c("a", "V1", "c", "V4"))
-test(2321.29, names(data.table(cbind(1), cbind(2))), c("V1", "V1"))
+test(2321.28, options=c(datatable.old.matrix.autoname=TRUE),
+     names(data.table(a=1, cbind(2), c=3, 4)), c("a", "V1", "c", "V4"))
+test(2321.29, options=c(datatable.old.matrix.autoname=TRUE),
+     names(data.table(cbind(1), cbind(2))), c("V1", "V1"))
 # also test behavior with a 0-column matrix
 M = cbind(1:3)
 test(2321.30, data.table(M[, 0L]), data.table(NULL))
 test(2321.31, data.table(a=1:3, M[, 0L]), data.table(a=1:3))
 
-local({
-  old = options(datatable.old.matrix.autoname=FALSE)
-  on.exit(options(old))
-  
-  test(2321.32, names(data.table(a=1, cbind(2), c=3, 4)), c("a", "V2", "c", "V4"))
-  # particularly buggy old behavior: can easily result in duplicate names
-  test(2321.33, names(data.table(cbind(1), cbind(2))), c("V1", "V2"))
-  M = cbind(1:3)
-  test(2321.34, data.table(M[, 0L]), data.table(NULL))
-  test(2321.35, data.table(a=1:3, M[, 0L]), data.table(a=1:3))
-
-  # a more subtle version of this as expressed in #5367
-  DT <- data.table(Counts=c(10, 20), Severity=c(1, 2))
-  test(2321.36, names(DT[,.(New_name = Severity %*% Counts)]), "New_name")
-})
+test(2321.32, names(data.table(a=1, cbind(2), c=3, 4)), c("a", "V2", "c", "V4"))
+# particularly buggy old behavior: can easily result in duplicate names
+test(2321.33, names(data.table(cbind(1), cbind(2))), c("V1", "V2"))
+M = cbind(1:3)
+test(2321.34, data.table(M[, 0L]), data.table(NULL))
+test(2321.35, data.table(a=1:3, M[, 0L]), data.table(a=1:3))
+
+# a more subtle version of this as expressed in #5367
+DT <- data.table(Counts=c(10, 20), Severity=c(1, 2))
+test(2321.36, names(DT[,.(New_name = Severity %*% Counts)]), "New_name")
 
 # New fctr() helper: like factor() but retaining order by default #4837
 test(2322.01, levels(fctr(c("b","a","c"))), c("b","a","c"))
diff --git a/man/data.table-options.Rd b/man/data.table-options.Rd
index 63c79fdf6..439e88ef2 100644
--- a/man/data.table-options.Rd
+++ b/man/data.table-options.Rd
@@ -110,7 +110,7 @@
 
 \section{Back-compatibility Options}{
   \describe{
-    \item{\code{datatable.old.matrix.autoname}}{Logical, default \code{TRUE}. Governs how the output of
+    \item{\code{datatable.old.matrix.autoname}}{Logical, default \code{FALSE}. Governs how the output of
       expressions like \code{data.table(x=1, cbind(1))} will be named. When \code{TRUE}, it will be named
       \code{V1}, otherwise it will be named \code{V2}.
     }