In reading through the Intel Consent Decree I’m reminded of both the years I spent interpreting and writing documents like this and how glad I am they are past me.
Legal agreements, as an unfortunate side effect of what they need to accomplish, are virtually impossible to interpret with a high degree of accuracy and this one is no exception.
However, we can make some reasonably reliable guesses about how it will work.
Let’s cover both why agreements of this type are misleading and the likely results that will come out of this one.
Why Contracts like Consent Decrees Are Misleading
The problem resides in the nature of this being a complex agreement between two parties with vastly different backgrounds and goals coupled with a third enforcement element that typically had no part in the creation process and yet must interpret the result.
In this case you have a contract that defines Intel behavior drafted by Intel who is motivated to limit the impact of the contract, reviewed and altered by a second party (FTC) who far removed from Intel that desires a specific behavior, and this contract is enforced by people in both entities (Compliance Officer/Technical Consultant/Political Appointees) who had nothing to do with creating it.
Often you can’t even be sure how the agreement is supposed to work because of the conflicts between the two initial parties let alone how the enforcement folks will interpret the result.
Ideally, to understand the impact of an agreement like this you should chat with the FTC appointed on-site resource who will be in charge of assuring enforcement (in this case a defined Technical Consultant reporting to the commission who will earn their $200K- per year salary).
And revisit this every time that person changes. If no Technical Consultant is selected (and this is only an option in this case), the FTC becomes the compliance officer and interpretation will be even less well defined. It would appear to be in both Intel and the FTC’s interest to have this roll staffed.
Consent decrees can have unforeseen consequences as well. For instance, the Microsoft Consent Decree resulted in a lot of domestic litigation, became the basis for European Commission’s actions resulting in fines and ended up making Microsoft more competitive by forcing them to better document their code and lead in interoperability. In that case the consent decree may have ultimately done more good than damage to the company.
In IBM’s case, that Consent Decree became an anchor around IBM’s neck, contributed significantly to the company’s decline, and eventually resulted in the creation of Microsoft who then passed IBM for dominance of the segment.
While it took time, both consent decrees fundamentally changed their respective companies but how the companies reacted to them vastly changed their respective outcomes.
These outcomes, particularly Microsoft’s, were not anticipated at the time the agreements were executed. This means that even the opinion of the appointed compliance officer (technical consultant) may not be particularly accurate long term.
Consent Decree Impact
I’m going to try to order these by level of impact not by agreement section. Be aware that many of these provisions, to be effective and timely, would require the use of the optional Technical Consultant:
The biggest impact is that companies that have a complaint from an Intel competitor about an alleged anti-competitive practice will have a fast track to the FTC to resolve this complaint. A process that could take over a decade is, in effect, shifted to one that could take weeks.
While Intel hasn’t admitted guilt this provision treats them as if they were guilty, depending on how it is implemented, could have the greatest impact. The role of Technical Consultant would be critical to quickly separating unfounded claims from real problems.
The agreement relatively tightly defines how Intel can discount removing some of the tools, like retroactive tiered discounting, that Intel was accused of using in an illegal fashion.
The pricing section is the most robust and appears to bracket Intel reasonably well with regard to how they can use their market size and financial resources as a weapon against competitors. There are some investment loopholes which appear but may not in hindsight be contained enough allowing for substantive financial investments to what is a very small pool of powerful customers.
The agreement limits Intel’s ability to make design changes that are simply targeted to hurt a competing product. Loophole is that unintentional damage or “bugs” are excluded (I’ve seen bugs defined as features) in this provision and time to correct the “bugs” is not specified but the burden of proof is on Intel to argue lack of intent which offsets this risk. In other words just because Intel calls it a “bug” doesn’t make it so. This is likely where a compliance officer would be asked to render an opinion.
One area that will likely be a source of contention, given what happened between the European Union their Special Master and Microsoft under a similar provision, is that if interoperability. The intent of these sections (II & VI) is to provide organizations like NVIDIA and the ATI side of AMD with the interface information they need to develop timely products like Graphics chips that have to interface closely with the processor. Language like “reasonable” can mean vastly different things to folks at Nvidia and Intel and even where defined conditions may cause Intel to believe they can behave in a way the commission or compliance officer believes is unreasonable. From the standpoint of the enforcement body section VI is poorly written and will likely be a source of considerable contention as a result.
The agreement provides special protections to VIA with respect to earlier litigation and lengthens some of those remedies and it backstops the AMD/Intel settlement with regard to Global Foundries.
The agreement requires Intel to disclose that benchmarks they publish on their own products may be unreliable which is kind of like the statement on the bottom of Cigarettes saying Smoking may be harmful to your health.
From Intel’s perspective this agreement likely has minor impact given much of what it asks Intel not to do they believe they already aren’t doing. From the FTC’s perspective it effectively prevents Intel from doing things that were illegal.
There is a lot of room for interpretation in the agreement suggesting that there will likely be a number of disagreements that will result and that until a viable Technical Consultant is selected coupled with an Intel Compliance Officer who has some authority and is willing to use it critical changes that may be required are at risk as is this agreement.
Given the existing settlement agreement with AMD which covers many similar provisions, the impact on AMD is more limited than it is on NVIDIA and Via both of whose positions are significantly improved for as long as the agreement is effectively enforced. As to Intel, the agreement could either critically damage or help them in a similar way to how the EU agreement eventually helped Microsoft.
It will depend on how quickly Intel shifts from gaming the agreement to use completive loopholes to embracing concepts like interoperability more aggressively to expand their available market. Intel’s biggest risk these days is ARM – and there AMD and Via are likely closer to allies than competitors anyway.
Rob Enderle is one of the last Inquiry Analysts. Inquiry Analysts are paid to stay up to date on current events and identify trends and either explain the trends or make suggestions, tactical and strategic, on how to best take advantage of them. Currently, he provides his services to most of the major technology and media companies. The opinions expressed in this commentary are solely those of the writer.