Agents & Hacking

Security and Alarm System

While sentinels provide security in the real world, agents provide security in the Matrix. Agents are able to issue orders to sentinels from within the Matrix (as seen in M1), and in The Matrix Online video game when the General (who commands red-eye agents according to Agent Pace) says, "Not all bees follow the hive" (sentinels were purposely given a "buzzing" sound in the movies in order to creep people out more during sentinel scenes). Zion is a threat to the machine world and is therefore a primary target of sentinels, which explains the Agents' high priority to capture Morpheus and obtain access codes to Zion's mainframe.

It is tempting to think that Agents themselves are sentinels jacked into the Matrix, but this cannot be since Agents' code appears in green, not gold (see the "Gold Code" section of Seraph). Therefore, while agents and sentinels are in very close contact with each other, they are not one and the same. Neo is able to "hack" the Agent Smith "software" in M1, but he would not have been able to do the same to Seraph.

Being in charge of security, the Agents' "alarm system" within the Matrix seems to consist of people who experience shock or surprise. Every time we see a person morph into an Agent, it is always preceded by the person being astonished in some way, such as being attacked, seeing Morpheus leap over the cab of a semi, seeing a woman disappear in a phone booth, having one's cell phone stolen by a running man for no apparent reason, or realizing people are literally climbing down the wetwall of a building. Since Agents don't have access to all Matrix information at all times like the Merovingian, Architect and Oracle do, Agents cannot know which people to take over unless they first narrow it down to people experiencing a specific kind of shock.

When shocked people morph into Agents, the system is killing two birds with one stone. Not only is the system addressing a security threat, the system is also taking care of a potential redpill (rejection). When the bum saw Trinity disappear from the phone booth, he was probably seriously questioning his reality. He would be very likely to start rejecting the Matrix and causing more errors, so why not use him against a threat? As the final result, we just have a bum who got depressed and committed suicide in front of a subway train. But, the train driver saw Neo and Agent Smith jump up to the ceiling, followed by Neo doing an additional backflip 6 feet into the air. Logically, Agent Smith next morphed into the subway driver.

Agents also black out the memories of those they morph into, provided the Agents accomplish their mission. There is a Matrix comic that ends with a redpill hacker being killed inside the Matrix by an Agent. After the Agent kills her, the man the Agent morphed into cannot remember how the girl died.

When Morpheus and Neo are walking through the Agent training program, Morpheus says, "...they [agents] are the gatekeepers. They are guarding all the doors. They are holding all the keys...". At this point, Morpheus does not know about backdoors of the Matrix, so Morpheus is not referring to the keys and doors of backdoors. The keys and doors that Morpheus refers to must therefore be the hard line connections the rebels use to get in and out of the Matrix.

This brings even more meaning to the idea that the agents provide "security" to the Matrix system. They are the software firewall of the Matrix. In Zion's early years, when Zion rebels were first learning how to hack the Matrix, they likely encountered agents in the form of computer code that made hacking a challenge. In order to hack oneself into the Matrix, they must first get past the agents' monitoring of hard lines. That is exactly what a firewall does.

Agents vs. Arch-Agents

Agent Smith is clearly the leader of the other agents, and for this reason, the Wachowskis gave him the title "Arch-Agent". The Arch-Agent commands the other Agents and Sentinels in the real world. Arch-Agents may also be faster and more powerful than normal Agents. For example, when Agent Smith fights Neo in the subway station, Agent Smith delivers hundreds of punches to Neo in M1 in a matter of seconds. However, I happen to think that normal agents can punch just as fast, considering they can move their torsos fast enough to dodge bullets. I don't believe that Smith's extra hatred for humanity and the Matrix is a result of his Arch-Agent status, either. This will be discussed more in the Smith pages.

When Agent Smith is destroyed by Neo, all of the Agents are upgraded, and Arch-Agent Smith is replaced by Arch-Agent Johnson, who fights Morpheus on the semi truck. Since we never hear the term "Arch-Agent" in any of the movies (this is a term one only gets by reading the movie scripts), I will simply refer to Arch-Agent Smith as an "Agent Smith" to avoid confusion.


In the freeway chase scene of M2, Agents morph into a couple of police officers, and immediately they identify "the exile" (the Keymaker) as their primary target. This is certainly surprising after watching M1 and how obsessively the Agents pursue Morpheus for access codes to Zion's mainframe. Why not go after Morpheus instead? I believe this clues us in to the three main purposes of Agents when it comes to security of the system, in order of importance:

  1. Kill exile programs;
  2. Eliminate the Zion threat; and
  3. Eliminate the threat of those who hack into the Matrix. (Goal no. 2 would accomplish this goal, which is why no. 2 is a higher priority.)

Killing exile programs is the Agents' top priority because exile programs are still fulfilling their system function when the system has determined they shouldn't. For example, if an improved Keymaker program is created, the exile Keymaker is going to interfere with the new Keymaker's function at best, and at worst, prevent the new one from functioning at all. A better example might be weather: if two programs can both control clouds and wind and they try to do so at the same time, the errors that result in the system are potentially far more catastrophic than a single redpill hacker running around in the Matrix, even if that hacker has access codes to the Zion mainframe. Strange cloud formations and wind in opposite directions simultaneously might happen, causing thousands of bluepills to question reality in the Matrix.

There is a better and beautifully concise way to summarize all of the above purposes: To debug the Matrix, or to fight all occurances of anomalies within the Matrix. This includes anomalies living in Zion, anomalies hacking the Matrix, and programs that make anomalous decisions to go into exile. All of these examples of anomaly result from choice, which agents surely hate. Since choice will always be a necessary rule of the Matrix, the Agents will never be rid of anomalies to battle.

Power of the Agents

Once Agents are clued into a security situation, they can do just about anything they wish to address the situation. Not only can they morph in place of people and black out their memories, they can change the floorplan of a building, as we see in M1, and they can even rearrange the entire landscape of a parcel of land, as we see in "Beyond" (The Animatrix). They can also manipulate the very RSIs of bluepills as we see in M1 when Smith causes Neo's lips to seal shut. Agents don't have this same power over redpills since they are not still hard-wired to the system like bluepills are.

Puppets Designed for Failure

We must inevitably ask the question, "Why can't machines possessing technology of the year 2700 prevent humans possessing technology of the year 2100 from hacking into the Matrix?" Very simply, machines allow hacking to occur. If no people from Zion can hack into the Matrix, how could anyone ever find and train the One? How could the One ever choose the right door if the One doesn't even believe that billions of humans are actually slaves in pods in the real world?

This is not to say that agents cut any slack to anomalies. They continually put forth their full efforts to prevent redpills from hacking, to kill them while inside the Matrix, to get access codes to Zion, to destroy hacking ships with sentinels, etc. The problem is, despite the great power agents have, they have some crippling weaknesses that doom Agents to constant failure:

Agents are given these limitations as a necessary "evil" of allowing choice within the Matrix. Every person needs to be offered a legitimate choice to accept or reject the Matrix, and it is not within the mathematical morality of machines to interfere with the choice process by killing anyone who is considering swallowing a red pill. After Neo swallows the red pill, Neo surely experiences shock that would alarm Agents. Agents must still allow Neo to escape.

The Agents' efforts could be considered a kind of "show" that the machines must put on in order for Zion rebels to believe they have escaped the system of control even though they are actually helping it. If the system did not try to stop the One, Zionists would be suspicious that the One is part of the system of control, which he technically is even though he's not aware of it until the very end of his life.

However, Agents are not insincere actors in this show - they are puppets. Agents do not help Zionists locate and train the One - they try to kill them. And Agents do not help the One reach the Source - they try to kill him. Whether or not Agents are aware of the purpose of the One, the fact remains that their purpose is designed to fail. If Agents actually succeed at their purpose of eliminating the One, eliminating hackers, and eliminating Zion, then the "cataclysmic crash" that the Architect speaks of will happen just as it eventually does if the One chooses the wrong door (more on this later in the Conversations: M2: Neo & Architect page).

Having said that, the Agentsí efforts to kill redpills does serve one purpose in the grand scheme of the system: it slows the population growth of Zion. Even though wiping out 250,000 people in Zion is mere routine to the machines, it's still easier to wipe out 250,000 than it is to wipe out 260,000.

Aren't the machines risking a lot by giving Agents permission to kill the One?

I will discuss this in more detail on the Neo's Powers: Back to Life page, but in a nutshell, the One cannot die in the Matrix. This system-authorized protection of The One reinforces further the fact that agents are "puppets designed for failure". Agents cannot kill the biggest redpill of them all! (See "Another Possible Motivation" section of Smith: Purpose for more on this.)

Exile Agents

After Neo kills Agent Smith in M1, Agent Brown and Agent Jones look at each other and run in opposite directions. There is a little more to this scene than a simple fear of Neo. Neo has proven himself to be able to destroy Arch-Agent Smith, and the logical conclusion for machines would then be that Neo can even more easily destroy Agent Brown and Agent Jones. These two agents must have become immediately useless to the system and therefore scheduled for deletion, probably eventually choosing exile since their purpose of fighting anomaly within the Matrix can still be carried out. Neo wasn't the only thing the agents had to fear.

The next time we see agents (the beginning of M2), it is probably also the next time Neo sees them as well, and sure enough, the agents are upgrades. Agent Brown and Agent Jones are either hiding in exile or already deleted.

Back to Matrix System