Most people who ask whether a news source is biased are asking about outlets with anchors and opinion pages. Almost nobody asks about the wires. AP and Reuters are the plumbing of American news — they supply the raw copy that hundreds of local papers, broadcast segments and aggregator feeds rewrite — and precisely because they sound neutral, their choices go unexamined.
On September 16, 2026, the Federal Reserve raised its benchmark interest rate for the first time in three years. Eighteen outlets in our coverage set wrote about it. AP and Reuters both did. So did Bloomberg, four separate times. Comparing those pieces is a clean test of what "neutral" actually means in practice, because the facts were not in dispute. Only the framing was.
What AP and Reuters did that fifteen other outlets didn't
Both wires led with the politics.
Reuters ran the decision under the headline "Warsh on collision course with Trump administration." AP reported that Fed Chair Kevin Warsh was "likely to side with financial markets over Trump." The Financial Times, rated center-right in our system, went a step further and cited economists urging the Fed to "defy Donald Trump with rate rise."
That was the entire list. Three outlets out of eighteen treated the decision as a story about a Fed chair acting against the explicit preference of the president who appointed him. The other fifteen treated it as a technical question about inflation data.
This is worth sitting with, because it inverts what people expect from wire copy. The conventional criticism of AP and Reuters is that they sand the edges off a story — that neutrality flattens conflict into "both sides say." Here the wires were the ones who named the conflict, and the specialist financial press was the one that smoothed it over.
Bloomberg published at least four pieces in this cluster: bond traders' conviction, Citi's "one and done" call, European stocks, and copper prices. None of them mention Trump or Warsh by name in connection with political pressure. That is not an error. It is a beat decision — Bloomberg's markets desk covers markets. But a reader who got the Fed decision only from Bloomberg would not know that the central bank's independence was the live question that day.
How can I compare the same news from different sources?
The method that works is not "read a left outlet and a right outlet." It is narrower than that: pick one event, list every outlet that covered it, and write down the first fact each one chose to lead with. Framing lives in selection, not adjectives.
Run that on the Fed decision and the pattern falls out immediately:
Reuters and AP led with Warsh versus Trump. Bloomberg led with bond traders — specifically, that markets put the hike probability near 94%. NPR led with consumers, opening on what a hike means for car loans and credit card balances, and was the only outlet in the group to put the three-year gap since the last hike in its lede. MarketWatch led with the fiscal consequence, noting the 10-year Treasury yield above 5% makes America's debt bill "a lot bigger" the longer it holds. CoinDesk led with an institutional-credibility argument nobody else made — that Warsh had been cornered by market expectations, and failing to deliver a hawkish message "risks destroying the central bank's credibility."
Every one of those pieces is accurate. None of them is the same story. That gap between "accurate" and "complete" is the thing a single-source news diet cannot show you, and it is the reason source count matters more than any single article. You can compare two outlets side by side on our compare page or read the full framing breakdown of the Fed decision with all eighteen sources laid out.
Where the two wires actually differ
AP and Reuters landed in the same place here, but not by the same route, and the difference is characteristic of each.
Reuters wrote the conflict as a trajectory — a "collision course," something in motion with an outcome still pending. AP wrote it as a prediction about behavior: Warsh is "likely to side with financial markets over Trump." Reuters framed the institution against the administration. AP framed a person choosing between two constituencies.
Neither is more neutral than the other. Reuters' version carries an implication that the collision is structural and unavoidable. AP's version carries an implication that Warsh had a genuine choice and made it. If you are building a picture of whether the Fed retains independence or merely exercised it once, those two framings point in different directions.
This is the kind of distinction that bias ratings and media bias charts cannot capture, and it is why our longer look at what Reuters' coverage record actually shows focuses on story selection rather than a position on a chart. AP and Reuters are both rated center. On this story they were also both correct, both early, and both pointed at the same underlying fact. They still told you different things.
What both wires missed
Even-handedness cuts both ways, so: the wires had their own omissions.
Citi chief US economist Andrew Hollenhorst was the clearest dissenting voice in the entire coverage set, saying he saw no "momentum in the data for further hikes" and that "it's hard to make a strong case for a rate hike in this environment." That quote appeared in one Bloomberg video piece. Meanwhile Nomura's conversion from dovish to hawkish — a bank changing its forecast to match consensus — got its own CNBC story. A dissent ran once; the consensus ran everywhere.
MarketWatch's connection between 5% Treasury yields and the federal debt load ran in exactly one place. And while financial outlets tracked gold near $4,300 and copper edging higher, none noted that Zcash was up roughly 130% over thirty days and trading near its 2016 highs while bitcoin sat below $76,000 at a three-and-a-half-week low. That is a specific signal about where speculative money moves ahead of a tightening cycle, and it appeared only on CoinDesk.
There is one more absence worth naming. In our source breakdown for this story, the coverage came from center and adjacent outlets — Reuters, AP, Bloomberg, Inc, NPR, CNBC, FT. Not a single left-rated or right-rated outlet in the set. A decision that raises the cost of car loans and mortgages for every household in the country generated no partisan coverage at all, while the second Fed story in our feed that week drew the full spectrum. That is bias by omission operating at the level of an entire ideological wing rather than a single outlet.
What news sources are most reliable?
The honest answer is that reliability is not a property of an outlet, it is a property of a reading habit.
AP and Reuters earned their reputation here. On the single most consequential framing question of the day — is this a data story or a power story — they got it right while most of the specialist press did not. But they also underweighted the dissent, skipped the fiscal math, and missed the capital-flow signal. Reading only the wires would have left you better informed than reading only Bloomberg, and still short of the whole picture.
The Fed ultimately voted unanimously to raise rates, defying the administration's pressure to cut. The outlets that ran no political framing on Wednesday will have to retrofit it. Watching who does, and how quietly, is the next round of this same exercise — and it is the same skill as comparing sources side by side on any story.